ArxivArchitecture.txt

If there are any errors
please Abort, and run `arxiv_required` for required package installation, and start again
Please wait while we phrase the requested information from global arxiv[arxiv.org] servers 
------------>
---------------------------->
------------------------------------------------------>
 
Listening for Sirens: Locating and Classifying Acoustic Alarms in City Scenes (Letizia Marchegiani - 11 October, 2018)
Lastly, we use the denoised signals to localise the acoustic source on the horizon plane, by regressing the direction of arrival of the sound through a CNN architecture. Our experimental evaluation shows an average classification rate of 94%, and a median absolute error on the localisation of 7.5° when operating on audio frames of 0.5s, and of 2.5° when operating on frames of 2.5s
Link: https://arxiv.org/abs/1810.04989
====================================================
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding (Jacob Devlin - 10 October, 2018)
As a result, the pre-trained BERT representations can be fine-tuned with just one additional output layer to create state-of-the-art models for a wide range of tasks, such as question answering and language inference, without substantial task-specific architecture modifications.
Link: https://arxiv.org/abs/1810.04805
====================================================
Deep Inertial Poser: Learning to Reconstruct Human Pose from Sparse Inertial Measurements in Real Time (Yinghao Huang - 10 October, 2018)
A bi-directional RNN architecture leverages past and future information that is available at training time. To evaluate our method, we recorded DIP-IMU, a dataset consisting of $10$ subjects wearing 17 IMUs for validation in $64$ sequences with $330\,000$ time instants; this constitutes the largest IMU dataset publicly available
Link: https://arxiv.org/abs/1810.04703
====================================================
Multimodal Speech Emotion Recognition Using Audio and Text (Seunghyun Yoon - 10 October, 2018)
This architecture analyzes speech data from the signal level to the language level, and it thus utilizes the information within the data more comprehensively than models that focus on audio features. Our proposed model outperforms previous state-of-the-art methods in assigning data to one of four emotion categories (i.e., angry, happy, sad and neutral) when the model is applied to the IEMOCAP dataset, as reflected by accuracies ranging from 68.8% to 71.8%.
Link: https://arxiv.org/abs/1810.04635
====================================================
Adding 32-bit Mode to the ACL2 Model of the x86 ISA (Alessandro Coglio - 9 October, 2018)
The ACL2 model of the x86 Instruction Set Architecture was built for the 64-bit mode of operation of the processor
Link: https://arxiv.org/abs/1810.04313
====================================================
Inter-Scanner Harmonization of High Angular Resolution DW-MRI using Null Space Deep Learning (Vishwesh Nath - 9 October, 2018)
To use these data, we propose a new network architecture, the null space deep network (NSDN), to simultaneously learn on traditional observed/truth pairs (e.g., MRI-histology voxels) along with repeated observations without a known truth (e.g., scan-rescan MRI). NSDN significantly improved absolute performance relative to histology by 3.87% over CSD and 1.42% over a recently proposed deep neural network approach. More-over, it improved reproducibility on the paired data by 21.19% over CSD and 10.09% over a recently proposed deep approach. Finally, NSDN improved gen-eralizability of the model to a third in vivo human scanner (which was not used in training) by 16.08% over CSD and 10.41% over a recently proposed deep learn-ing approach
Link: https://arxiv.org/abs/1810.04260
====================================================
Geometry meets semantics for semi-supervised monocular depth estimation (Pierluigi Zama Ramirez - 9 October, 2018)
For inference, state-of-the-art encoder-decoder architectures for monocular depth estimation rely on effective feature representations learned at training time
Link: https://arxiv.org/abs/1810.04093
====================================================
A Distributed Reinforcement Learning Solution With Knowledge Transfer Capability for A Bike Rebalancing Problem (Ian Xiao - 9 October, 2018)
Motivated by such problem and the lack of smart autonomous solutions in this area, this project explored a new RL architecture called Distributed RL (DiRL) with Transfer Learning (TL) capability. DiRL achieved a 350% improvement in bike rebalancing autonomously and TL offered a 62.4% performance boost in managing an entire bike network
Link: https://arxiv.org/abs/1810.04058
====================================================
Deep Geodesic Learning for Segmentation and Anatomical Landmarking (Neslisah Torosdagli - 6 October, 2018)
In step 1, we propose a deep neu- ral network architecture with carefully designed regularization, and network hyper-parameters to perform image segmentation without the need for data augmentation and complex post- processing refinement. In step 2, we formulate the landmark localization problem directly on the geodesic space for sparsely- spaced anatomical landmarks. In step 3, we propose to use a long short-term memory (LSTM) network to identify closely- spaced landmarks, which is rather difficult to obtain using other standard detection networks. We used a very challenging CBCT dataset of 50 patients with a high-degree of craniomaxillofacial (CMF) variability that is realistic in clinical practice. Complementary to the quantitative analysis, the qualitative visual inspection was conducted for distinct CBCT scans from 250 patients with high anatomical variability
Link: https://arxiv.org/abs/1810.04021
====================================================
Comparison of U-net-based Convolutional Neural Networks for Liver Segmentation in CT (Hans Meine - 9 October, 2018)
Using a set of 219 liver CT datasets with reference segmentations from liver surgery planning, we evaluate the performance of several neural network classifiers based on 2D and 3D U-net architectures. An interesting observation is that slice-wise approaches perform surprisingly well, with mean and median Dice coefficients above 0.97, and may be preferable over 3D approaches given current hardware and software limitations.
Link: https://arxiv.org/abs/1810.04017
====================================================
Glioma Segmentation with Cascaded Unet (Dmitry Lachinov - 9 October, 2018)
Similar to recent methods for object detection, our implementation is based on neural networks; we propose modifications to the 3D UNet architecture and augmentation strategy to efficiently handle multimodal MRI input, besides this we introduce approach to enhance segmentation quality with context obtained from models of the same topology operating on downscaled data. We evaluate presented approach on BraTS 2018 dataset and discuss results.
Link: https://arxiv.org/abs/1810.04008
====================================================
Towards Two-Dimensional Sequence to Sequence Model in Neural Machine Translation (Parnia Bahar - 9 October, 2018)
This work investigates an alternative model for neural machine translation (NMT) and proposes a novel architecture, where we employ a multi-dimensional long short-term memory (MDLSTM) for translation modeling. Our proposed topology shows consistent improvements over attention-based sequence to sequence model on two WMT 2017 tasks, German$\leftrightarrow$English.
Link: https://arxiv.org/abs/1810.03975
====================================================
Rate-Accuracy Trade-Off In Video Classification With Deep Convolutional Neural Networks (Mohammad Jubran - 27 September, 2018)
Based on three CNN architectures and two action recognition datasets, we achieve 11%-94% saving in bitrate with marginal effect on classification accuracy. A model-based selection between multiple CNNs increases these savings further, to the point where, if up to 7% loss of accuracy can be tolerated, video classification can take place with as little as 3 kbps for the transport of the required compressed video information to the system implementing the CNN models.
Link: https://arxiv.org/abs/1810.03964
====================================================
Conditional Generative Refinement Adversarial Networks for Unbalanced Medical Image Semantic Segmentation (Mina Rezaei - 9 October, 2018)
The proposed architecture shows state-of-the-art results on LiTS-2017 for liver lesion segmentation, and two microscopic cell segmentation datasets MDA231, PhC-HeLa
Link: https://arxiv.org/abs/1810.03871
====================================================
Bootstrapped CNNs for Building Segmentation on RGB-D Aerial Imagery (Clint Sebastian - 8 October, 2018)
We train a detection algorithm from RGB-D images to obtain a segmented mask by using the CNN architecture DenseNet.First, we improve the performance of the model by applying a statistical re-sampling technique called Bootstrapping and demonstrate that more informative examples are retained. Second, the proposed method outperforms the non-bootstrapped version by utilizing only one-sixth of the original training data and it obtains a precision-recall break-even of 95.10% on our aerial imagery dataset.
Link: https://arxiv.org/abs/1810.03570
====================================================
NSGA-NET: A Multi-Objective Genetic Algorithm for Neural Architecture Search (Zhichao Lu - 8 October, 2018)
Experimental results suggest that combining the objectives of minimizing both an error metric and computational complexity, as measured by FLOPS, allows NSGA-Net to find competitive neural architectures near the Pareto front of both objectives on two different tasks, object classification and object alignment. NSGA-Net obtains networks that achieve 3.72% (at 4.5 million FLOP) error on CIFAR-10 classification and 8.64% (at 26.6 million FLOP) error on the CMU-Car alignment task
Link: https://arxiv.org/abs/1810.03522
====================================================
A Vertical PRF Architecture for Microblog Search (Flávio Martins - 8 October, 2018)
Thus, the proposed query expansion method uses a distributed search architecture and resource selection algorithms to provide an efficient query expansion process. Experiments on the TREC Microblog datasets show that the proposed approach can match or outperform standard PRF in MAP and NDCG@30, with a computational cost that is three orders of magnitude lower.
Link: https://arxiv.org/abs/1810.03519
====================================================
Multilingual sequence-to-sequence speech recognition: architecture, transfer learning, and language modeling (Jaejin Cho - 4 October, 2018)
We also explore different architectures for improving the prior multilingual seq2seq model. Experimental results show that the transfer learning approach from the multilingual model shows substantial gains over monolingual models across all 4 BABEL languages
Link: https://arxiv.org/abs/1810.03459
====================================================
Local Explanation Methods for Deep Neural Networks Lack Sensitivity to Parameter Values (Julius Adebayo - 8 October, 2018)
Our conjecture is that this phenomenon occurs because these explanations are dominated by the lower level features of a DNN, and that a DNN's architecture provides a strong prior which significantly affects the representations learned at these lower layers. NOTE: This work is now subsumed by our recent manuscript, Sanity Checks for Saliency Maps (to appear NIPS 2018), where we expand on findings and address concerns raised in Sundararajan et al.
Link: https://arxiv.org/abs/1810.03307
====================================================
Light-Weight RefineNet for Real-Time Semantic Segmentation (Vladimir Nekrasov - 8 October, 2018)
In particular, we adapt a powerful semantic segmentation architecture, called RefineNet, into the more compact one, suitable even for tasks requiring real-time performance on high-resolution inputs. Our fastest model undergoes a significant speed-up boost from 20 FPS to 55 FPS on a generic GPU card on 512x512 inputs with solid 81.1% mean iou performance on the test set of PASCAL VOC, while our slowest model with 32 FPS (from original 17 FPS) shows 82.7% mean iou on the same dataset. Alternatively, we showcase that our approach is easily mixable with light-weight classification networks: we attain 79.2% mean iou on PASCAL VOC using a model that contains only 3.3M parameters and performs only 9.3B floating point operations.
Link: https://arxiv.org/abs/1810.03272
====================================================
Triple Attention Mixed Link Network for Single Image Super Resolution (Xi Cheng - 7 October, 2018)
However, existing architectures have limitations due to the less sophisticated structure along with less strong representational power. In this work, to significantly enhance the feature representation, we proposed Triple Attention mixed link Network (TAN) which consists of 1) three different aspects (i.e., kernel, spatial and channel) of attention mechanisms and 2) fu-sion of both powerful residual and dense connections (i.e., mixed link)
Link: https://arxiv.org/abs/1810.03254
====================================================
DeepGeo: Photo Localization with Deep Neural Network (Sudharshan Suresh - 6 October, 2018)
A deep neural network based on the ResNet architecture is trained, and four different strategies of incorporating low-level cardinality information are presented. This model achieves an accuracy 20 times better than chance on a test dataset, which rises to 71.87% when taking the best of top-5 guesses. The network also beats human subjects in 4 out of 5 rounds of GeoGuessr.
Link: https://arxiv.org/abs/1810.03077
====================================================
Deep convolutional Gaussian processes (Kenneth Blomqvist - 6 October, 2018)
We propose deep convolutional Gaussian processes, a deep Gaussian process architecture with convolutional structure. In particular, we improve CIFAR-10 accuracy by over 10 percentage points.
Link: https://arxiv.org/abs/1810.03052
====================================================
Understanding Recurrent Neural Architectures by Analyzing and Synthesizing Long Distance Dependencies in Benchmark Sequential Datasets (Abhijit Mahalunkar - 6 October, 2018)
At present, the state-of-the-art computational models across a range of sequential data processing tasks, including language modeling, are based on recurrent neural network architectures. Finally, we demonstrate how understanding the characteristics of the LDDs in a dataset can inform better hyper-parameter selection for current state-of-the-art recurrent neural architectures and also aid in understanding them...
Link: https://arxiv.org/abs/1810.02966
====================================================
Towards Self-Tuning Parameter Servers (Chris Liu - 6 October, 2018)
Parameter Server (PS) is a popular system architecture for large-scale machine learning systems; and by self-tuning we mean while a long-running ML job is iteratively training the expert-suggested model, the system is also iteratively learning which system setting is more efficient for that job and applies it online. Experiments show that our techniques can reduce the completion times of a variety of long-running TensorFlow jobs from 1.4x to 18x.
Link: https://arxiv.org/abs/1810.02935
====================================================
RCCNet: An Efficient Convolutional Neural Network for Histological Routine Colon Cancer Nuclei Classification (Shabbeer Basha S H - 30 September, 2018)
This paper proposes an efficient Convolutional Neural Network (CNN) based architecture for classification of histological routine colon cancer nuclei named as RCCNet. The proposed method has achieved a classification accuracy of 80.61% and 0.7887 weighted average F1 score
Link: https://arxiv.org/abs/1810.02797
====================================================
Robotics CTF (RCTF), a playground for robot hacking (Gorka Olalde Mendia - 1 October, 2018)
We describe the architecture of the RCTF and provide 9 scenarios where hackers can challenge the security of different robotic setups
Link: https://arxiv.org/abs/1810.02690
====================================================
A Comparative Survey of Optical Wireless Technologies: Architectures and Applications (Mostafa Zaman Chowdhury - 5 October, 2018)
We survey the key technologies for understanding OWC and present state-of-the-art criteria in aspects, such as classification, spectrum use, architecture, and applications
Link: https://arxiv.org/abs/1810.02594
====================================================
Weisfeiler and Leman Go Neural: Higher-order Graph Neural Networks (Christopher Morris - 4 October, 2018)
In recent years, graph neural networks (GNNs) have emerged as a powerful neural architecture to learn vector representations of nodes and graphs in a supervised, end-to-end fashion. The following work investigates GNNs from a theoretical point of view and relates them to the $1$-dimensional Weisfeiler-Leman graph isomorphism heuristic ($1$-WL). We show that GNNs have the same expressiveness as the $1$-WL in terms of distinguishing non-isomorphic (sub-)graphs
Link: https://arxiv.org/abs/1810.02244
====================================================
Learning Finer-class Networks for Universal Representations (Julien Girard - 4 October, 2018)
We show that our method learns more universal representations than state-of-the-art, leading to significantly better results on 10 target-tasks from multiple domains, using several network architectures, either alone or combined with networks learned at a coarser semantic level.
Link: https://arxiv.org/abs/1810.02126
====================================================
Improving the Segmentation of Anatomical Structures in Chest Radiographs using U-Net with an ImageNet Pre-trained Encoder (Maayan Frid-Adar - 4 October, 2018)
In this paper we investigate the latest fully-convolutional architectures for the task of multi-class segmentation of the lungs field, heart and clavicles in a chest radiograph. We evaluate all models on a common benchmark of 247 X-ray images from the JSRT database and ground-truth segmentation masks from the SCR dataset. This model outperformed the current state-of-the-art methods tested on the same benchmark, with Jaccard overlap scores of 96.1% for lung fields, 90.6% for heart and 85.5% for clavicles.
Link: https://arxiv.org/abs/1810.02113
====================================================
Towards Fast and Energy-Efficient Binarized Neural Network Inference on FPGA (Cheng Fu - 4 October, 2018)
By analyzing local properties of images and the learned BNN kernel weights, we observe an average of $\sim$78% input similarity and $\sim$59% weight similarity among weight kernels, measured by our proposed metric in common network architectures
Link: https://arxiv.org/abs/1810.02068
====================================================
Domain Specific Approximation for Object Detection (Ting-Wu Chin - 3 October, 2018)
category-aware image size scaling and proposals scaling, for two state-of-the-art deep learning-based object detection meta-architectures. up to 7.5x speedup for dynamic domain-specific approximation
Link: https://arxiv.org/abs/1810.02010
====================================================
Understanding Weight Normalized Deep Neural Networks with Rectified Linear Units (Yixi Xu - 3 October, 2018)
In particular, for an $L_{1,\infty}$ weight normalized network, the approximation error can be controlled by the $L_1$ norm of the output layer, and the corresponding generalization error only depends on the architecture by the square root of the depth.
Link: https://arxiv.org/abs/1810.01877
====================================================
A Deep Learning Architecture for De-identification of Patient Notes: Implementation and Evaluation (Kaung Khin - 2 October, 2018)
We test this architecture on two gold standard datasets and show that the architecture achieves state-of-the-art performance on both data sets while also converging faster than other systems without the use of dictionaries or other knowledge sources.
Link: https://arxiv.org/abs/1810.01570
====================================================
On Self Modulation for Generative Adversarial Networks (Ting Chen - 2 October, 2018)
We propose and study an architectural modification, self-modulation, which improves GAN performance across different data sets, architectures, losses, regularizers, and hyperparameter settings. In a large-scale empirical study we observe a relative decrease of $5\%-35\%$ in FID. Furthermore, all else being equal, adding this modification to the generator leads to improved performance in $124/144$ ($86\%$) of the studied settings
Link: https://arxiv.org/abs/1810.01365
====================================================
FutureGAN: Anticipating the Future Frames of Video Sequences using Spatio-Temporal 3d Convolutions in Progressively Growing Autoencoder GANs (Sandra Aigner - 2 October, 2018)
Our approach extends the recently introduced progressive growing of GANs (PGGAN) architecture by Karras et al. [18]. We already achieve promising results for frame resolutions of 128 x 128 px over a variety of datasets ranging from synthetic to natural frame sequences, while theoretically not being limited to a specific frame resolution
Link: https://arxiv.org/abs/1810.01325
====================================================
Findings of the E2E NLG Challenge (Ondřej Dušek - 2 October, 2018)
We compare 62 systems submitted by 17 institutions, covering a wide range of approaches, including machine learning architectures -- with the majority implementing sequence-to-sequence models (seq2seq) -- as well as systems based on grammatical rules and templates.
Link: https://arxiv.org/abs/1810.01170
====================================================
Training compact deep learning models for video classification using circulant matrices (Alexandre Araujo - 8 October, 2018)
We propose very compact models for video classification based on state-of-the-art network architectures such as Deep Bag-of-Frames, NetVLAD and NetFisherVectors
Link: https://arxiv.org/abs/1810.01140
====================================================
Performance Comparison of some Synchronous Adders (P Balasubramanian - 2 October, 2018)
The adder architectures mentioned were implemented by targeting a typical case PVT specification (high threshold voltage, supply voltage of 1.05V and operating temperature of 25 degrees Celsius) of the Synopsys 32/28nm CMOS technology
Link: https://arxiv.org/abs/1810.01115
====================================================
NU-LiteNet: Mobile Landmark Recognition using Convolutional Neural Networks (Chakkrit Termritthikun - 2 October, 2018)
This paper describes the development of the architecture of a new convolutional neural network model, NU-LiteNet. The model size of NU-LiteNet is therefore 2.6 times smaller than that of SqueezeNet
Link: https://arxiv.org/abs/1810.01074
====================================================
The Profiling Machine: Active Generalization over Knowledge (Filip Ilievski - 1 October, 2018)
We describe two generic state-of-the-art neural architectures that can be easily instantiated as profiling machines to generate expectations and applied to any kind of knowledge to fill gaps
Link: https://arxiv.org/abs/1810.00782
====================================================
Procedural Noise Adversarial Examples for Black-Box Attacks on Deep Neural Networks (Kenneth T. Co - 30 September, 2018)
We show that it is possible to construct practical black-box attacks with low computational cost against robust neural network architectures such as Inception v3 and Inception ResNet v2 on the ImageNet dataset. Perlin noise attacks achieve at least 90% top 1 error across all classifiers. More worryingly, we show that most Perlin noise perturbations are "universal" in that they generalize, as adversarial examples, across large portions of the dataset, with up to 73% of images misclassified using a single perturbation
Link: https://arxiv.org/abs/1810.00470
====================================================
Pseudo-Random Number Generation using Generative Adversarial Networks (Marcello De Bernardi - 30 September, 2018)
Furthermore, we showcase a number of interesting modifications to the standard GAN architecture. At best, subjected to the NIST test suite, the trained generator passed around 99% of test instances and 98% of overall tests, outperforming a number of standard non-cryptographic PRNGs.
Link: https://arxiv.org/abs/1810.00378
====================================================
Neural Entity Reasoner for Global Consistency in NER (Xiaoxiao Yin - 30 September, 2018)
2) the specific interaction-pooling mechanism, allowing it to connect each local word to multiple global entities, and 3) the deep architecture, allowing it to bootstrap the recognized entity set from coarse to fine
Link: https://arxiv.org/abs/1810.00347
====================================================
Resource Management in Fog/Edge Computing: A Survey (Cheol-Ho Hong - 29 September, 2018)
This article reviews publications as early as 1991, with 85% of the publications between 2013-2018, to identify and classify the architectures, infrastructure, and underlying algorithms for managing resources in fog/edge computing.
Link: https://arxiv.org/abs/1810.00305
====================================================
NICE: Noise Injection and Clamping Estimation for Neural Network Quantization (Chaim Baskin - 2 October, 2018)
This leads to state-of-the-art results on various regression and classification tasks, e.g., ImageNet classification with architectures such as ResNet-18/34/50 with low as 3-bit weights and activations
Link: https://arxiv.org/abs/1810.00162
====================================================
Optimization of Circuits for IBM's five-qubit Quantum Computers (Gerhard W. Dueck - 28 September, 2018)
Two architectures with five qubits, one with 16, and one with 20 qubits are available to run experiments. In this paper, we show how Clifford+T circuits can efficiently be mapped into the two IBM quantum computers with 5 qubits
Link: https://arxiv.org/abs/1810.00129
====================================================
Audio-Visual Speech Recognition With A Hybrid CTC/Attention Architecture (Stavros Petridis - 28 September, 2018)
To the best of our knowledge, this is the first time that such a hybrid architecture architecture is used for audio-visual recognition of speech. We use the LRS2 database and show that the proposed audio-visual model leads to an 1.3% absolute decrease in word error rate over the audio-only model and achieves the new state-of-the-art performance on LRS2 database (7% word error rate). We also observe that the audio-visual model significantly outperforms the audio-based model (up to 32.9% absolute improvement in word error rate) for several different types of noise as the signal-to-noise ratio decreases.
Link: https://arxiv.org/abs/1810.00108
====================================================
Celer Network: Bring Internet Scale to Every Blockchain (Mo Dong - 28 September, 2018)
Celer Network embraces a layered architecture with clean abstractions that enable rapid evolution of each individual component, including a generalized state channel and sidechain suite that supports fast and generic off-chain state transitions; a provably optimal value transfer routing mechanism that achieves an order of magnitude higher throughput compared to state-of-the-art solutions; a powerful development framework and runtime for off-chain applications; and a new cryptoeconomic model that provides network effect, stable liquidity, and high availability for the off-chain ecosystem.
Link: https://arxiv.org/abs/1810.00037
====================================================
Universal and Dynamic Locally Repairable Codes with Maximal Recoverability via Sum-Rank Codes (Umberto Martínez-Peñas - 28 September, 2018)
Furthermore, the local linear codes (thus the localities, local distances and local fields) can be efficiently and dynamically modified without global recoding or changes in architecture or outer code, while preserving MR, easily adapting to new hot and cold data. Reed-Solomon codes with local replication and Cartesian products are recovered from the given construction when $ r=1 $ and $ h = 0 $, respectively
Link: https://arxiv.org/abs/1809.11158
====================================================
Semantic Segmentation for Urban Planning Maps based on U-Net (Zhiling Guo - 30 September, 2018)
In this study, we select urban planning maps as a representative sample and investigate the feasibility of utilizing U-shape fully convolutional based architecture to perform end-to-end map semantic segmentation. The experimental results obtained from the test area in Shibuya district, Tokyo, demonstrate that our proposed method could achieve a very high Jaccard similarity coefficient of 93.63% and an overall accuracy of 99.36%
Link: https://arxiv.org/abs/1809.10862
====================================================
Adaptive Input Representations for Neural Language Modeling (Alexei Baevski - 30 September, 2018)
We perform a systematic comparison of popular choices for a self-attentional architecture. We achieve a new state of the art on the WikiText-103 benchmark of 20.51 perplexity, improving the next best known result by 8.7 perplexity. On the Billion word benchmark, we achieve a state of the art of 24.14 perplexity.
Link: https://arxiv.org/abs/1809.10853
====================================================
Using Multi-task and Transfer Learning to Solve Working Memory Tasks (T. S. Jayram - 28 September, 2018)
We propose a new architecture called Memory-Augmented Encoder-Solver (MAES) that enables transfer learning to solve complex working memory tasks adapted from cognitive psychology. We show by extensive experimentation that the trained MAES models achieve task-size generalization, i.e., they are capable of handling sequential inputs 50 times longer than seen during training, with appropriately large memory modules
Link: https://arxiv.org/abs/1809.10847
====================================================
Patient Risk Assessment and Warning Symptom Detection Using Deep Attention-Based Neural Networks (Ivan Girardi - 27 September, 2018)
We use an attention-based convolutional neural network architecture trained on 600,000 doctor notes in German. These approaches achieve 79% and 66% precision, respectively, but on a confidence threshold of 0.6, precision increases to 85% and 75%, respectively
Link: https://arxiv.org/abs/1809.10804
====================================================
nnU-Net: Self-adapting Framework for U-Net-Based Medical Image Segmentation (Fabian Isensee - 27 September, 2018)
The adaptation of the U-Net to novel problems, however, comprises several degrees of freedom regarding the exact architecture, preprocessing, training and inference. At the time of manuscript submission, nnU-Net achieves the highest mean dice scores across all classes and seven phase 1 tasks (except class 1 in BrainTumour) in the online leaderboard of the challenge.
Link: https://arxiv.org/abs/1809.10486
====================================================
Image Reconstruction Using Deep Learning (Po-Yu Liu - 27 September, 2018)
The architecture incorporates a hybrid of convolutional and deconvolutional layers along with symmetric connections. The denoising network achieved statistically significant 0.38dB, 0.68dB, and 1.04dB average PSNR gains over benchmark traditional algorithms in experiments with image peak values 4, 2, and 1
Link: https://arxiv.org/abs/1809.10410
====================================================
Semantically Invariant Text-to-Image Generation (Shagan Sah - 26 September, 2018)
Our work ties these concepts together by creating an architecture that can enable bidirectional generation of images and text. Qualitative and quantitative evaluations demonstrate that MMVR improves upon existing text conditioned image generation results by over 20%, while integrating visual and text modalities.
Link: https://arxiv.org/abs/1809.10274
====================================================
Deeply Informed Neural Sampling for Robot Motion Planning (Ahmed H. Qureshi - 26 September, 2018)
DeepSMP's neural architecture comprises of a Contractive AutoEncoder which encodes given workspaces directly from a raw point cloud data, and a Dropout-based stochastic deep feedforward neural network which takes the workspace encoding, start and goal configuration, and iteratively generates feasible samples for SMPs to compute end-to-end collision-free optimal paths. The results show that on average our method is at least 7 times faster in point-mass and rigid-body case and about 28 times faster in 6-link robot case than the existing state-of-the-art.
Link: https://arxiv.org/abs/1809.10252
====================================================
High Performance Zero-Memory Overhead Direct Convolutions (Jiyuan Zhang - 19 September, 2018)
In this paper, we demonstrate that direct convolution, when implemented correctly, eliminates all memory overhead, and yields performance that is between 10% to 400% times better than existing high performance implementations of convolution layers on conventional and embedded CPU architectures
Link: https://arxiv.org/abs/1809.10170
====================================================
Developmental Bayesian Optimization of Black-Box with Visual Similarity-Based Transfer Learning (Amaury Depierre - 3 October, 2018)
This architecture allows a robot to optimize autonomously hyper-parameters that need to be tuned from any action and/or vision module, treated as a black-box. As example, the system has been used to optimized 9 continuous hyper-parameters of a professional software (Kamido) both in simulation and with a real robot (industrial robotic arm Fanuc) with a total of 13 different objects. The robot is able to find a good object-specific optimization in 68 (simulation) or 40 (real) trials. Moreover, with the real robot, we show that the method consistently outperforms the manual optimization from an expert with less than 2 hours of training time to achieve more than 88% of success.
Link: https://arxiv.org/abs/1809.10141
====================================================
Learning Navigation Behaviors End to End (Hao-Tien Lewis Chiang - 26 September, 2018)
We train these policies in small, static environments with Shaped-DDPG, an adaptation of the Deep Deterministic Policy Gradient (DDPG) reinforcement learning method which optimizes reward and network architecture. Over 500 meters of on-robot experiments show , these policies generalize to new environments and moving obstacles, are robust to sensor, actuator, and localization noise, and can serve as robust building blocks for larger navigation tasks. The path following and point and point policies are 83% and 56% more successful than the baseline, respectively.
Link: https://arxiv.org/abs/1809.10124
====================================================
Satellite Imagery Multiscale Rapid Detection with Windowed Networks (Adam Van Etten - 24 September, 2018)
airplanes versus airports) we find that using two different detectors at different scales is very effective with negligible runtime cost.We evaluate large test images at native resolution and find mAP scores of 0.2 to 0.8 for vehicle localization, with the YOLT architecture achieving both the highest mAP and fastest inference speed.
Link: https://arxiv.org/abs/1809.09978
====================================================
Public-Safety LTE: Communication Services, Standardization Status, and Disaster-Resilient Architecture (Zeeshan Kaleem - 26 September, 2018)
Simulation results verified that with DR-PSLTE architecture, delay is reduced by 20% as compared with the conventional centralized computing architecture.
Link: https://arxiv.org/abs/1809.09617
====================================================
Low Complexity Full Duplex MIMO: Novel Analog Cancellation Architectures and Transceiver Design (George C. Alexandropoulos - 22 September, 2018)
In this paper, we present two novel architectures for the analog canceller comprising of reduced number of cancellation elements, compared to the state of the art, and simple multiplexers for efficient signal routing among the transmit and receive radio frequency chains
Link: https://arxiv.org/abs/1809.09474
====================================================
An Efficient Framework for Implementing Persist Data Structures on Remote NVM (Teng Ma - 25 September, 2018)
In this architecture, the NVM devices can be shared by multiple servers and provide recoverable persistent data structures. Specifically, thanks to operation batching, local memory caching and efficient concurrency control, the throughput of operations on eight widely used data structures is improved by 6$\sim$22 $\times$ without lowering the consistency promising.
Link: https://arxiv.org/abs/1809.09395
====================================================
Adaptive Resonant Beam Charging for Intelligent Wireless Power Transfer (Qingqing Zhang - 25 September, 2018)
Due to its open-loop architecture, RBC faces the challenge of providing dynamic current and voltage to optimize battery charging performance. Numerical evaluation demonstrates that ARBC can save 61% battery charging energy and 53%-60% supplied energy compared with RBC
Link: https://arxiv.org/abs/1809.09364
====================================================
Software for Sparse Tensor Decomposition on Emerging Computing Architectures (Eric Phipps - 24 September, 2018)
We show that we are competitive with state-of-the-art approaches available in the literature while having the advantage of being able to run on a wider of variety of architectures with a single code.
Link: https://arxiv.org/abs/1809.09175
====================================================
On Reinforcement Learning for Full-length Game of StarCraft (Zhen-Jia Pang - 23 September, 2018)
The reinforcement training algorithm for this architecture is also investigated. On a 64x64 map and using restrictive units, we achieve a winning rate of more than 99\% against the difficulty level-1 built-in AI. Through the curriculum transfer learning algorithm and a mixture of combat model, we can achieve over 93\% winning rate of Protoss against the most difficult non-cheating built-in AI (level-7) of Terran, training within two days using a single machine with only 48 CPU cores and 8 K40 GPUs
Link: https://arxiv.org/abs/1809.09095
====================================================
Deep Confidence: A Computationally Efficient Framework for Calculating Reliable Errors for Deep Neural Networks (Isidro Cortes-Ciriano - 24 September, 2018)
Deep learning architectures have proved versatile in a number of drug discovery applications, including the modelling of in vitro compound activity. Using a set of 24 diverse IC50 data sets from ChEMBL 23, we show that Snapshot Ensembles perform on par with Random Forest (RF) and ensembles of independently trained deep neural networks
Link: https://arxiv.org/abs/1809.09060
====================================================
Internet of NanoThings: Concepts and Applications (Ebtesam Almazrouei - 20 September, 2018)
Section 2 describes Internet of NanoThing(IoNT), its network architecture, and the challenges of nanoscale communication which is essential for enabling IoNT. Section 3 gives some practical applications of IoNT
Link: https://arxiv.org/abs/1809.08914
====================================================
Person Identification using Seismic Signals generated from Footfalls (Bodhibrata Mukhopadhyay - 24 September, 2018)
This paper presents a Fog computing architecture for implementing footfall based biometric system using widespread geographically distributed geophones (vibration sensor). We have tested our biometric system on an indigenous database (created by us) containing 46000 footfall events from 8 individuals and achieved an accuracy of 73%, 90% and 95% in case of 1, 5 and 10 footsteps per sample. DS8BP compresses the original footfall events (sampled at 8 kHz) by a factor of 108 and also acts as a smoothing filter
Link: https://arxiv.org/abs/1809.08783
====================================================
Image Denoising and Super-Resolution using Residual Learning of Deep Convolutional Network (Rohit Pardasani - 21 September, 2018)
Our model nearly replicates the architecture of existing state-of-the-art deep learning models for super-resolution and denoising
Link: https://arxiv.org/abs/1809.08229
====================================================
SLIDER: Fast and Efficient Computation of Banded Sequence Alignment (Mohammed Alser - 18 September, 2018)
The second key idea is to design a hardware accelerator design that adopts modern FPGA (field-programmable gate array) architectures to fur-ther boost the performance of our algorithm. The addition of SLIDER as a pre-alignment step reduces the execution time of five state-of-the-art sequence align-ers by up to 18.8x
Link: https://arxiv.org/abs/1809.07858
====================================================
Design and Implementation of High-throughput PCIe with DMA Architecture between FPGA and PowerPC (Kun Cheng - 17 September, 2018)
The DMA architecture based on FPGA is compatible with the Xilinx PCIe core while the DMA architecture based on POWERPC is compatible with VxBus of VxWorks. The design is compatible with Xilinx FPGA Kintex Ultrascale Family, and operates with the Xilinx PCIe endpoint Generation 1 with lane configurations x8. A data throughput of more than 666 MBytes/s(memory write with data from FPGA to PowerPC) has been achieved with the single PCIe Gen1 x8 lanes endpoint of this design, PowerPC and FPGA can send memory write request to each other.
Link: https://arxiv.org/abs/1809.07702
====================================================
A Microbenchmark Characterization of the Emu Chick (Jeffrey Young - 7 September, 2018)
Our results demonstrate that for many basic operations the Emu Chick can use available memory bandwidth more efficiently than a more traditional, cache-based architecture although bandwidth usage suffers for computationally intensive workloads like SpMV. Moreover, the Emu Chick provides stable, predictable performance with up to 65% of the peak bandwidth utilization on a random-access pointer chasing benchmark with weak locality.
Link: https://arxiv.org/abs/1809.07696
====================================================
OxIOD: The Dataset for Deep Inertial Odometry (Changhao Chen - 20 September, 2018)
However, the lack of sufficient labelled data for training and testing various architectures limits the proliferation of adopting DNNs in IMU-based tasks. Our dataset contains 158 sequences totalling more than 42 km in total distance, much larger than previous inertial datasets
Link: https://arxiv.org/abs/1809.07491
====================================================
Faster Training of Mask R-CNN by Focusing on Instance Boundaries (Roland S. Zimmermann - 3 October, 2018)
While the computational costs are increased slightly, the increment is negligible considering the high computational cost of the Mask R-CNN architecture. In a default Mask R-CNN setup, we achieve a training speed up of 29% and an overall improvement of 8.1% on the MS COCO metrics compared to the baseline.
Link: https://arxiv.org/abs/1809.07069
====================================================
Exploring Visual Relationship for Image Captioning (Ting Yao - 19 September, 2018)
Specifically, we present Graph Convolutional Networks plus Long Short-Term Memory (dubbed as GCN-LSTM) architecture that novelly integrates both semantic and spatial object relationships into image encoder. More remarkably, GCN-LSTM increases CIDEr-D performance from 120.1% to 128.7% on COCO testing set.
Link: https://arxiv.org/abs/1809.07041
====================================================
Security and Protocol Exploit Analysis of the 5G Specifications (Roger Piqueras Jover - 20 September, 2018)
Moreover, the entire 5G security architecture relies on the assumption of impractical carrier and roaming agreements and the management of public keys from all global operators. The comparison with 4G LTE protocol exploits reveals that the 5G security specifications, as of Release 15, do not fully address the user privacy and network availability concerns, where one edge case can compromise the privacy, security and availability of 5G users and services.
Link: https://arxiv.org/abs/1809.06925
====================================================
3D segmentation of mandible from multisectional CT scans by convolutional neural networks (Bingjiang Qiu - 18 September, 2018)
The proposed convolutional neural network adopts the architecture of the U-Net and then combines the resulting 2D segmentations from three different planes into a 3D segmentation. We implement such a segmentation approach on 11 neck CT scans and then evaluate the performance. We achieve an average dice coefficient of $ 0.89 $ on two testing mandible segmentation
Link: https://arxiv.org/abs/1809.06752
====================================================
SECS: Efficient Deep Stream Processing via Class Skew Dichotomy (Boyuan Feng - 7 September, 2018)
Finally, we design a systematic framework, SECS, to dynamically detect class skew, processing interpretation and compilation, as well as select the most accurate architectures under the runtime resource budget. Extensive evaluations show that SECS can realize end-to-end classification speedups by a factor of 3x to 11x relative to state-of-the-art convolutional neural networks, at a higher accuracy.
Link: https://arxiv.org/abs/1809.06691
====================================================
GritNet 2: Real-Time Student Performance Prediction with Domain Adaptation (Byung-Hak Kim - 7 September, 2018)
In particular, we first review recently-developed GritNet architecture which is the current state of the art for student performance prediction problem, and introduce a new unsupervised domain adaptation method to transfer a GritNet trained on a past course to a new course without any (students' outcome) label
Link: https://arxiv.org/abs/1809.06686
====================================================
MBS: Macroblock Scaling for CNN Model Reduction (Yu-Hsun Lin - 18 September, 2018)
Our proposed macroblock scaling (MBS) algorithm can be applied to various CNN architectures to reduce their model size. These applicable models range from compact CNN models such as MobileNet (25.53% reduction, ImageNet) and ShuffleNet (20.74% reduction, ImageNet) to ultra-deep ones such as ResNet-101 (51.67% reduction, ImageNet) and ResNet-1202 (72.71% reduction, CIFAR-10) with negligible accuracy degradation
Link: https://arxiv.org/abs/1809.06569
====================================================
Scattering Networks for Hybrid Representation Learning (Edouard Oyallon - 17 September, 2018)
Indeed, using hybrid architectures, we achieve the best results with predefined representations to-date, while being competitive with end-to-end learned CNNs. Specifically, even applying a shallow cascade of small-windowed scattering coefficients followed by 1$\times$1-convolutions results in AlexNet accuracy on the ILSVRC2012 classification task. Moreover, by combining scattering networks with deep residual networks, we achieve a single-crop top-5 error of 11.4% on ILSVRC2012
Link: https://arxiv.org/abs/1809.06367
====================================================
Binary Classification of Alzheimer Disease using sMRI Imaging modality and Deep Learning (Ahsan Bin Tufail - 8 September, 2018)
In this study, by utilizing the concept of transfer learning in deep learning, we propose a classification framework to differentiate subjects with Clinical Dementia Rating (CDR) of zero from subjects with CDR greater than zero by using deep learning architectures such as Xception and Inception version 3 in the Keras deep learning library. The attained validation set accuracies are as high as 99.12% for the Inception version 3 network and 97.97% for the Xception network
Link: https://arxiv.org/abs/1809.06209
====================================================
DASNet: Reducing Pixel-level Annotations for Instance and Semantic Segmentation (Chuang Niu - 17 September, 2018)
Our architecture, named DASNet, consists of three modules: detection, attention, and segmentation. Our method demonstrates substantially improved performance compared to existing semi-supervised approaches on PASCAL VOC 2012 dataset.
Link: https://arxiv.org/abs/1809.06013
====================================================
FermiNets: Learning generative machines to generate efficient neural networks via generative synthesis (Alexander Wong - 16 September, 2018)
To tackle this challenge, we explore the following idea: Can we learn generative machines to automatically generate deep neural networks with efficient network architectures? In this study, we introduce the idea of generative synthesis, which is premised on the intricate interplay between a generator-inquisitor pair that work in tandem to garner insights and learn to generate highly efficient deep neural networks that best satisfies operational requirements. Experimental results for image classification, semantic segmentation, and object detection tasks illustrate the efficacy of generative synthesis in producing generators that automatically generate highly efficient deep neural networks (which we nickname FermiNets) with higher model efficiency and lower computational costs (reaching >10x more efficient and fewer multiply-accumulate operations than several tested state-of-the-art networks), as well as higher energy efficiency (reaching >4x improvements in image inferences per joule consumed on a Nvidia Tegra X2 mobile processor)
Link: https://arxiv.org/abs/1809.05989
====================================================
A Distributed Learning Architecture for Scientific Imaging Problems (A. Panousopoulou - 27 September, 2018)
We apply the resulting, Spark-compliant, architecture on two emerging use cases from the scientific imaging domain, namely: (a) the space variant deconvolution of galaxy imaging surveys (astrophysics), (b) the super-resolution based on coupled dictionary training (remote sensing). We conduct evaluation studies considering relevant datasets, and the results report at least 60\% improvement in time response against the conventional computing solutions
Link: https://arxiv.org/abs/1809.05956
====================================================
Comparison of Deep Learning and the Classical Machine Learning Algorithm for the Malware Detection (Mohit Sewak - 16 September, 2018)
We studied the performance of the classical RF and DNN with 2, 4 & 7 layers architectures with the four different feature sets, and found that irrespective of the features inputs, the classical RF accuracy outperforms the DNN.
Link: https://arxiv.org/abs/1809.05889
====================================================
An investigation of a deep learning based malware detection system (Mohit Sewak - 16 September, 2018)
In the investigation, we experiment with different combination of Deep Learning architectures including Auto-Encoders, and Deep Neural Networks with varying layers over Malicia malware dataset on which earlier studies have obtained an accuracy of (98%) with an acceptable False Positive Rates (1.07%). In our proposed approach, besides improving the previous best results (99.21% accuracy and a False Positive Rate of 0.19%) indicates that Deep Learning based systems could deliver an effective defense against malware
Link: https://arxiv.org/abs/1809.05888
====================================================
Accident Forecasting in CCTV Traffic Camera Videos (Ankit Shah - 15 September, 2018)
Finally, we demonstrate the performance of accident forecasting in our dataset using Faster R-CNN and an Accident LSTM architecture. We achieved an average of 1.359 seconds in terms of Time-To-Accident measure with an Average Precision of 47.36 %
Link: https://arxiv.org/abs/1809.05782
====================================================
OffsetNet: Deep Learning for Localization in the Lung using Rendered Images (Jake Sganga - 15 September, 2018)
In this paper, we introduce a deep learning architecture, called OffsetNet, to accurately localize a bronchoscope in the lung in real-time. After training on only 30 minutes of recorded camera images in conserved regions of a lung phantom, OffsetNet tracks the bronchoscope's motion on a held-out recording through these same regions at an update rate of 47 Hz and an average position error of 1.4 mm. After training on simulated images, OffsetNet tracks the bronchoscope's motion in less conserved regions at an average position error of 2.4 mm, which meets conservative thresholds required for successful tracking.
Link: https://arxiv.org/abs/1809.05645
====================================================
Ground Truth for training OCR engines on historical documents in German Fraktur and Early Modern Latin (Uwe Springmann - 14 September, 2018)
The special form of GT as line image/transcription pairs makes it directly usable to train state-of-the-art recognition models for OCR software employing recurring neural networks in LSTM architecture such as Tesseract 4 or OCRopus. We also provide some pretrained OCRopus models for subcorpora of our dataset yielding between 95\% (early printings) and 98\% (19th century Fraktur printings) character accuracy rates on unseen test cases, a Perl script to harmonize GT produced by different transcription rules, and give hints on how to construct GT for OCR purposes which has requirements that may differ from linguistically motivated transcriptions.
Link: https://arxiv.org/abs/1809.05501
====================================================
Multi-Kernel Diffusion CNNs for Graph-Based Learning on Point Clouds (Lasse Hansen - 14 September, 2018)
They are predestined to overcome certain limitations of conventional grid-based architectures and will enable efficient handling of point clouds or related graphical data representations, e.g. We validated our approach for learning point descriptors as well as semantic classification on real 3D point clouds of human poses and demonstrate an improvement from 85% to 95% in Dice overlap with our multi-kernel approach.
Link: https://arxiv.org/abs/1809.05370
====================================================
Deep CNN Frame Interpolation with Lessons Learned from Natural Language Processing (Kian Ghodoussi - 16 September, 2018)
From there, we demonstrate the effectiveness of our approach by presenting novel deep CNN frame interpolation architecture that is comparable to the state of the art interpolation models with a fraction of the complexity.
Link: https://arxiv.org/abs/1809.05286
====================================================
Full Workspace Generation of Serial-link Manipulators by Deep Learning based Jacobian Estimation (Peiyuan Liao - 13 September, 2018)
The architecture consists of two neural networks: an estimation net that approximates the manipulator Jacobian, and a confidence net that measures the confidence of the approximation. We also introduce M3 (Manipulability Maps of Manipulators), a MATLAB robotics library based on [2](RTB), the datasets generated by which are used by this work. Implementations of the algorithm (based on Keras[3]), including benchmark evaluation script, are available at https://github.com/liaopeiyuan/Jacobian-Estimation 
Link: https://arxiv.org/abs/1809.05020
====================================================
Deep learning to achieve clinically applicable segmentation of head and neck anatomy for radiotherapy (Stanislav Nikolov - 12 September, 2018)
Adopting a deep learning approach, we demonstrate a 3D U-Net architecture that achieves performance similar to experts in delineating a wide range of head and neck OARs. The model was trained on a dataset of 663 deidentified computed tomography (CT) scans acquired in routine clinical practice and segmented according to consensus OAR definitions. We demonstrate its generalisability through application to an independent test set of 24 CT scans available from The Cancer Imaging Archive collected at multiple international sites previously unseen to the model, each segmented by two independent experts and consisting of 21 OARs commonly segmented in clinical practice
Link: https://arxiv.org/abs/1809.04430
====================================================
Deep learning for time series classification: a review (Hassan Ismail Fawaz - 12 September, 2018)
In this article, we study the current state of the art performance of deep learning algorithms for TSC by presenting an empirical study of the most recent DNN architectures for TSC. We also provide an open source deep learning framework to the TSC community where we implemented each of the compared approaches and evaluated them on a univariate TSC benchmark (the UCR archive) and 12 multivariate time series datasets. By training 8,730 deep learning models on 97 time series datasets, we propose the most exhaustive study of DNNs for TSC to date.
Link: https://arxiv.org/abs/1809.04356
====================================================
Rapid Training of Very Large Ensembles of Diverse Neural Networks (Abdul Wasay - 12 September, 2018)
In particular, our approach trains an ensemble of $100$ variants of deep neural networks with diverse architectures up to $6 \times$ faster as compared to existing approaches
Link: https://arxiv.org/abs/1809.04270
====================================================
Searching for Efficient Multi-Scale Architectures for Dense Image Prediction (Liang-Chieh Chen - 11 September, 2018)
The design of neural network architectures is an important component for achieving state-of-the-art performance with machine learning systems across a broad array of tasks. Based on a survey of techniques in dense image prediction, we construct a recursive search space and demonstrate that even with efficient random search, we can identify architectures that outperform human-invented architectures and achieve state-of-the-art performance on three dense prediction tasks including 82.7\% on Cityscapes (street scene parsing), 71.3\% on PASCAL-Person-Part (person-part segmentation), and 87.9\% on PASCAL VOC 2012 (semantic image segmentation). Additionally, the resulting architecture is more computationally efficient, requiring half the parameters and half the computational cost as previous state of the art systems.
Link: https://arxiv.org/abs/1809.04184
====================================================
On-Demand TDMA for Energy Efficient Data Collection with LoRa and Wake-up Receiver (Rajeev Piyare - 11 September, 2018)
To overcome this drawback, we propose a heterogeneous network architecture and an energy-efficient On-demand TDMA communication scheme improving both the device lifetime and the data latency of standard LoRa networks. Experimental results show a data reliability of 100% and a round-trip latency on the order of milliseconds with end devices dissipating less than 46 mJ when active and 1.83 μW during periods of inactivity, lasting up to 3 years on a 1200 mAh Lithium battery.
Link: https://arxiv.org/abs/1809.04142
====================================================
Parallel Separable 3D Convolution for Video and Volumetric Data Understanding (Felix Gonda - 11 September, 2018)
Lastly, we empirically show that PmSCn is applicable to different backbone architectures, such as ResNet, DenseNet, and UNet, for different applications, such as video action recognition, MRI brain segmentation, and electron microscopy segmentation. In all three applications, we replace the 3D convolution layers in state-of-the art models with PmSCn and achieve around 14% improvement in test performance and 40% reduction in model size and on average.
Link: https://arxiv.org/abs/1809.04096
====================================================
On The Alignment Problem In Multi-Head Attention-Based Neural Machine Translation (Tamer Alkhouli - 11 September, 2018)
This work investigates the alignment problem in state-of-the-art multi-head attention models based on the transformer architecture. Using the proposed approach, we achieve up to $3.8$ % BLEU improvement when using the dictionary, in comparison to $2.4$ % BLEU in the baseline case. We also propose alignment pruning to speed up decoding in alignment-based neural machine translation (ANMT), which speeds up translation by a factor of $1.8$ without loss in translation performance. We carry out experiments on the shared WMT 2016 English$\to$Romanian news task and the BOLT Chinese$\to$English discussion forum task.
Link: https://arxiv.org/abs/1809.03985
====================================================
5G Massive MIMO Architectures: Self-Backhauled Small Cells versus Direct Access (Andrea Bonfante - 11 September, 2018)
In this paper, we focus on one of the key technologies for the fifth-generation wireless communication networks, massive multiple-input-multiple-output (mMIMO), by investigating two of its most relevant architectures: 1) to provide in-band wireless backhauling to a dense deployment of self-backhauled small cells (SCs) acting as communication relays to end-users, and 2) to provide direct wireless access (DA) to end-users. We find that the ad-hoc deployment of self-backhauled SCs closer to the users (UEs) with optimal resource partition and with directive antenna patterns, provides rate improvements for cell-edge UEs that amount to 30%, and a tenfold gain as compared to mMIMO DA architecture with pilot reuse 3 and reuse 1, respectively
Link: https://arxiv.org/abs/1809.03953
====================================================
UAV Aided Aerial-Ground IoT for Air Quality Sensing in Smart City: Architecture, Technologies and Implementation (Zhiwen Hu - 11 September, 2018)
The architecture of this system consists of four layers: the sensing layer to collect data, the transmission layer to enable bidirectional communications, the processing layer to analyze and process the data, and the presentation layer to provide graphic interface for users. Our implementation has been deployed in Peking University and Xidian University since February 2018, and has collected about 100 thousand effective data samples by June 2018.
Link: https://arxiv.org/abs/1809.03746
====================================================
Jointly Learning to See, Ask, and GuessWhat (Aashish Venkatesh - 10 September, 2018)
We show that the introduction of our new architecture combined with these learning regimes yields an increase of 19.5% in task success accuracy with respect to a baseline model that treats submodules independently. With this increase, we reach an accuracy comparable to state-of-the-art models that use reinforcement learning, with the advantage that our architecture is entirely differentiable and thus easier to train
Link: https://arxiv.org/abs/1809.03408
====================================================
A Comparison of CNN-based Face and Head Detectors for Real-Time Video Surveillance Applications (Le Thanh Nguyen-Meidine - 10 September, 2018)
This paper compares the accuracy and complexity of state-of-the-art CNN architectures that are suitable for face and head detection
Link: https://arxiv.org/abs/1809.03336
====================================================
Towards JointUD: Part-of-speech Tagging and Lemmatization using Recurrent Neural Networks (Gor Arakelyan - 10 September, 2018)
The results demonstrate the viability of the proposed multitask architecture, although its performance still remains far from state-of-the-art.
Link: https://arxiv.org/abs/1809.03211
====================================================
Shallow vs deep learning architectures for white matter lesion segmentation in the early stages of multiple sclerosis (Francesco La Rosa - 10 September, 2018)
In this work, we present a comparison of a shallow and a deep learning architecture for the automated segmentation of white matter lesions in MR images of multiple sclerosis patients. All methods were trained on 32 patients, and the evaluation was performed on a pure test set of 73 cases. Results show low lesion-wise false positives (30%) for the deep learning architecture, whereas the shallow architecture yields the best Dice coefficient (63%) and volume difference (19%). Combining both shallow and deep architectures further improves the lesion-wise metrics (69% and 26% lesion-wise true and false positive rate, respectively).
Link: https://arxiv.org/abs/1809.03185
====================================================
SHOMA at Parseme Shared Task on Automatic Identification of VMWEs: Neural Multiword Expression Tagging with High Generalisation (Shiva Taslimipoor - 9 September, 2018)
We employ a neural architecture comprising of convolutional and recurrent layers with the addition of an optional CRF layer at the top. It outperformed all participating systems in both open and closed tracks with the overall macro-average MWE-based F1 score of 58.09 averaged among all languages
Link: https://arxiv.org/abs/1809.03056
====================================================
A Neural Temporal Model for Human Motion Prediction (Anand Gopalakrishnan - 14 September, 2018)
Key aspects of our proposed system include: 1) a novel, two-level processing architecture that aids in generating planned trajectories, 2) a simple set of easily computable features that integrate simple derivative information into the model, and 3) a novel multi-objective loss function that helps the model to slowly progress from the simpler task of next-step prediction to the harder task of multi-step closed-loop prediction
Link: https://arxiv.org/abs/1809.03036
====================================================
Accelerating Viterbi Algorithm using Custom Instruction Approach (Waqar Ahmad - 8 September, 2018)
In this paper, we propose to utilize the custom instruction approach to efficiently implement the widely used Viterbi decoding algorithm by adding the assembly language instructions to the ISA of DLX, PicoJava II and NIOS II processors, which represent RISC, stack and FPGA-based soft-core processor architectures, respectively. By using the custom instruction approach, the execution time of the Viterbi algorithm is significantly improved by approximately 3 times for DLX and PicoJava II, and by 2 times for NIOS II.
Link: https://arxiv.org/abs/1809.02887
====================================================
A Supervised Learning Methodology for Real-Time Disguised Face Recognition in the Wild (Saumya Kumaar - 8 September, 2018)
Along the same lines, we propose a deep learning architecture for disguised facial recognition (DFR). The algorithm put forward in this paper detects 20 facial key-points in the first stage, using a 14-layered convolutional neural network (CNN). Our key-point feature prediction accuracy is 65% while the classification rate is 72.4%. Moreover, the architecture works at 19 FPS, thereby performing in almost real-time
Link: https://arxiv.org/abs/1809.02875
====================================================
Adversarial Learning for Image Forensics Deep Matching with Atrous Convolution (Yaqi Liu - 8 September, 2018)
In DMAC, atrous convolution is adopted to extract features with rich spatial information, the correlation layer based on the skip architecture is proposed to capture hierarchical features, and atrous spatial pyramid pooling is constructed to localize tampered regions at multiple scales. Extensive experiments, conducted on 21 generated testing sets and two public datasets, demonstrate the effectiveness of the proposed framework and the superior performance of DMAC.
Link: https://arxiv.org/abs/1809.02791
====================================================
Metamorphic Relation Based Adversarial Attacks on Differentiable Neural Computer (Alvin Chan - 7 September, 2018)
The unique architecture of DNC contributes to its state-of-the-art performance in tasks which requires the ability to represent variables and data structure as well as to store data over long timescales
Link: https://arxiv.org/abs/1809.02444
====================================================
Predicting Lung Nodule Malignancies by Combining Deep Convolutional Neural Network and Handcrafted Features (Shulong Li - 7 September, 2018)
We then trained 3D CNNs modified from three state-of-the-art 2D CNN architectures (AlexNet, VGG-16 Net and Multi-crop Net) to extract the CNN features learned at the output layer. For each 3D CNN, the CNN features combined with the 29 handcrafted features were used as the input for the support vector machine (SVM) coupled with the sequential forward feature selection (SFS) method to select the optimal feature subset and construct the classifiers. The patient cohort includes 431 malignant nodules and 795 benign nodules extracted from the LIDC/IDRI database
Link: https://arxiv.org/abs/1809.02333
====================================================
Cell-aware Stacked LSTMs for Modeling Sentences (Jihun Choi - 6 September, 2018)
We dub this architecture Cell-aware Stacked LSTM (CAS-LSTM) and show from experiments that our models achieve state-of-the-art results on benchmark datasets for natural language inference, paraphrase detection, and sentiment classification.
Link: https://arxiv.org/abs/1809.02279
====================================================
ProdSumNet: reducing model parameters in deep neural networks via product-of-sums matrix decompositions (Chai Wah Wu - 6 September, 2018)
As an example, by using this decomposition on a reference CNN architecture for MNIST with over 3x10^6 trainable parameters, we are able to obtain an accuracy of 98.44% using only 3554 trainable parameters.
Link: https://arxiv.org/abs/1809.02209
====================================================
A Benchmarking of DCM Based Architectures for Position and Velocity Controlled Walking of Humanoid Robots (Giulio Romualdi - 6 September, 2018)
We show which implementation of the above control architecture allows the robot to achieve a walking velocity of 0.41 meters per second.
Link: https://arxiv.org/abs/1809.02167
====================================================
Panoptic Segmentation with a Joint Semantic and Instance Segmentation Network (Daan de Geus - 6 September, 2018)
For instance segmentation, a Mask R-CNN type of architecture is used, while the semantic segmentation branch is augmented with a Pyramid Pooling Module. Results for this method are submitted to the COCO and Mapillary Joint Recognition Challenge 2018. Our approach achieves a PQ score of 17.6 on the Mapillary Vistas validation set and 27.2 on the COCO test-dev set.
Link: https://arxiv.org/abs/1809.02110
====================================================
Named Entity Recognition on Noisy Data using Images and Text (1-page abstract) (Diego Esteves - 3 September, 2018)
In this paper, we propose a novel multi-level architecture that does not rely on any specific linguistic resource or encoded rule. Experimental tests against state-of-the-art NER for Twitter on the Ritter dataset present competitive results (0.59 F-measure), indicating that this approach may lead towards better NER models.
Link: https://arxiv.org/abs/1809.01964
====================================================
Efficient Egocentric Visual Perception Combining Eye-tracking, a Software Retina and Deep Learning (Nina Hristozova - 5 September, 2018)
We present ongoing work to harness biological approaches to achieving highly efficient egocentric perception by combining the space-variant imaging architecture of the mammalian retina with Deep Learning methods. By pre-processing images collected by means of eye-tracking glasses to control the fixation locations of a software retina model, we demonstrate that we can reduce the input to a DCNN by a factor of 3, reduce the required number of training epochs and obtain over 98% classification rates when training and validating the system on a database of over 26,000 images of 9 object classes.
Link: https://arxiv.org/abs/1809.01633
====================================================
Bimodal network architectures for automatic generation of image annotation from text (Mehdi Moradi - 5 September, 2018)
There is a clear advantage obtained from the architecture with pre-trained imaging network. The centroids of the ROIs marked by this network were on average at a distance equivalent to 5.1% of the image width from the centroids of the ground truth ROIs.
Link: https://arxiv.org/abs/1809.01610
====================================================
CNNs-based Acoustic Scene Classification using Multi-Spectrogram Fusion and Label Expansions (Weiping Zheng - 5 September, 2018)
In the framework, a single CNN architecture is applied onto multiple spectrograms for feature extraction. Specifically, accuracies of 0.9744, 0.8865 and 0.7778 are obtained for the LITIS Rouen dataset, the DCASE Development set and Evaluation set respectively.
Link: https://arxiv.org/abs/1809.01543
====================================================
Image Manipulation with Perceptual Discriminators (Diana Sungatullina - 5 September, 2018)
We demonstrate the merits of the new architecture in a series of qualitative and quantitative comparisons with baseline approaches and state-of-the-art frameworks for unaligned image translation.
Link: https://arxiv.org/abs/1809.01396
====================================================
Retinal Vessel Segmentation under Extreme Low Annotation: A Generative Adversarial Network Approach (Avisek Lahiri - 5 September, 2018)
The proposed method is an extension of our previous work with the addition of a new unsupervised adversarial loss and a structured prediction based architecture. We experiment with extreme low annotation budget (0.8 - 1.6% of contemporary annotation size)
Link: https://arxiv.org/abs/1809.01348
====================================================
Unsupervised Statistical Machine Translation (Mikel Artetxe - 4 September, 2018)
Our method profits from the modular architecture of SMT: we first induce a phrase table from monolingual corpora through cross-lingual embedding mappings, combine it with an n-gram language model, and fine-tune hyperparameters through an unsupervised MERT variant. In addition, iterative backtranslation improves results further, yielding, for instance, 14.08 and 26.22 BLEU points in WMT 2014 English-German and English-French, respectively, an improvement of more than 7-10 BLEU points over previous unsupervised systems, and closing the gap with supervised SMT (Moses trained on Europarl) down to 2-5 BLEU points
Link: https://arxiv.org/abs/1809.01272
====================================================
Decentralized Search on Decentralized Web (Ziliang Lai - 18 August, 2018)
DWeb also clicks well with future Internet architectures, such as Named Data Networking (NDN).Search engines have been an inseparable element of the Web. Contemporary ("Web 2.0") search engines, however, provide centralized services
Link: https://arxiv.org/abs/1809.00939
====================================================
Automated Instruction Stream Throughput Prediction for Intel and AMD Microarchitectures (Jan Laukemann - 10 October, 2018)
We present the Open Source Architecture Code Analyzer (OSACA), a static analysis tool for predicting the execution time of sequential loops comprising x86 instructions under the assumption of an infinite first-level cache and perfect out-of-order scheduling
Link: https://arxiv.org/abs/1809.00912
====================================================
Image Reassembly Combining Deep Learning and Shortest Path Problem (M. -M. Paumard - 4 September, 2018)
The main contributions of this work are: 1) several deep neural architectures to predict the relative position of image fragments that outperform the previous state of the art; 2) casting the reassembly problem into the shortest path in a graph problem for which we provide several construction algorithms depending on available information; 3) a new dataset of images taken from the Metropolitan Museum of Art (MET) dedicated to image reassembly for which we provide a clear setup and a strong baseline.
Link: https://arxiv.org/abs/1809.00898
====================================================
Texar: A Modularized, Versatile, and Extensible Toolkit for Text Generation (Zhiting Hu - 4 September, 2018)
In Texar, model architecture, losses, and learning processes are fully decomposed. Texar is released under Apache license 2.0 at https://github.com/asyml/texar.
Link: https://arxiv.org/abs/1809.00794
====================================================
Programmable Memristive Threshold Logic Gate Array (Olga Krestinskaya - 2 September, 2018)
The proposed TLG array operation does not depend on input signal and time pulses, comparing to the existing architectures. The on-chip area and power dissipation of the simulated $3\times 4$ TLG array is $1463 μm^2$ and $425 μW$, respectively.
Link: https://arxiv.org/abs/1809.00419
====================================================
On the Role of Event Boundaries in Egocentric Activity Recognition from Photostreams (Alejandro Cartas - 6 September, 2018)
Furthermore, we collected a new annotated dataset acquired by 15 people by a wearable photo-camera and we used it to show the generalization capabilities of several deep learning based architectures to unseen users.
Link: https://arxiv.org/abs/1809.00402
====================================================
Learning to Navigate Autonomously in Outdoor Environments : MAVNet (Saumya Kumaar - 2 September, 2018)
Based on the Inception-v3 architecture, our system performs better in terms of processing complexity and accuracy than many existing models for imitation learning. With the extensive amount of drone data that we collected, we have been able to navigate successfully through roads without crashing or overshooting, with an accuracy of 98.44%
Link: https://arxiv.org/abs/1809.00396
====================================================
Evaluation of Neural Networks for Image Recognition Applications: Designing a 0-1 MILP Model of a CNN to create adversarials (Lucas Schelkes - 1 September, 2018)
We follow up on (Fischetti & Jo, December, 2017) and show how standard convolutional neural network can be optimized to a more sophisticated capsule architecture. 2. 3
Link: https://arxiv.org/abs/1809.00216
====================================================
3D Segmentation with Exponential Logarithmic Loss for Highly Unbalanced Object Sizes (Ken C. L. Wong - 24 September, 2018)
By combining skip connections and deep supervision with respect to the computational feasibility of 3D segmentation, we propose a fast converging and computationally efficient network architecture for accurate segmentation. We achieve an average Dice coefficient of 82% on brain segmentation with 20 labels, with the ratio of the smallest to largest object sizes as 0.14%. Less than 100 epochs are required to reach such accuracy, and segmenting a 128x128x128 volume only takes around 0.4 s.
Link: https://arxiv.org/abs/1809.00076
====================================================
Bioinspired Straight Walking Task-Space Planner (Carlo Tiseo - 31 August, 2018)
The data show that the proposed architecture can generate behaviour in line with human walking strategies for both the CoM and the foot swing. Moreover, being the planner able to generate a single stride in less than 140 ms and sequences of 10 strides in less than 600 ms, it allows an online task-space planning for locomotion
Link: https://arxiv.org/abs/1808.10799
====================================================
Ensemble Sequence Level Training for Multimodal MT: OSU-Baidu WMT18 Multimodal Machine Translation System Report (Renjie Zheng - 31 August, 2018)
Our systems ensemble several models using different architectures and training methods and achieve the best performance for three subtasks: En-De and En-Cs in task 1 and (En+De+Fr)-Cs task 1B.
Link: https://arxiv.org/abs/1808.10592
====================================================
Multi-Cell Multi-Task Convolutional Neural Networks for Diabetic Retinopathy Grading (Kang Zhou - 11 October, 2018)
Considering the resolution of retinal image is very high, where small pathological tissues can be detected only with large resolution image and large local receptive field are required to identify those late stage disease, but directly training a neural network with very deep architecture and high resolution image is both time computational expensive and difficult because of gradient vanishing/exploding problem, we propose a \textbf{Multi-Cell} architecture which gradually increases the depth of deep neural network and the resolution of input image, which both boosts the training time but also improves the classification accuracy. Experimental results on the Kaggle dataset show that our method achieves a Kappa of 0.841 on test set which is the 4-th rank of all state-of-the-arts methods
Link: https://arxiv.org/abs/1808.10564
====================================================
Total Recall: Understanding Traffic Signs using Deep Hierarchical Convolutional Neural Networks (Sourajit Saha - 30 August, 2018)
In this paper, we propose a novel and one-for-all architecture that aces multiple benchmarks with better overall score than the state-of-the-art architectures. With this we score 99.33% Accuracy in German sign recognition benchmark and 99.17% Accuracy in Belgian traffic sign classification benchmark
Link: https://arxiv.org/abs/1808.10524
====================================================
Contribution of Glottal Waveform in Speech Emotion: A Comparative Pairwise Investigation (Zhongzhe Xiao - 30 August, 2018)
In experiments of generation of a performance-driven hierarchical classifier architecture, and pairwise classification on individual emotional states, the low difference between accuracies obtained from speech signal and glottal signal proved that a majority of emotional cues in speech could be conveyed through glottal waveform. The best distinguishable emotional pair by glottal waveform is intense anger against moderate sadness, with the accuracy of 92.45%
Link: https://arxiv.org/abs/1808.10144
====================================================
The Impact of Preprocessing on Deep Representations for Iris Recognition on Unconstrained Environments (Luiz A. Zanlorensi - 29 August, 2018)
In this context, we propose the use of deep representations, more specifically, architectures based on VGG and ResNet-50 networks, for dealing with the images using (and not) iris segmentation and normalization. Our results show that the approach using non-normalized and only circle-delimited iris images reaches a new state of the art in the official protocol of the NICE.II competition, a subset of the UBIRIS database, one of the most challenging databases on unconstrained environments, reporting an average Equal Error Rate (EER) of 13.98% which represents an absolute reduction of about 5%.
Link: https://arxiv.org/abs/1808.10032
====================================================
MemComputing Integer Linear Programming (Fabio L. Traversa - 29 August, 2018)
We first describe a new circuit architecture of memcomputing machines specifically designed to solve for the linear inequalities representing a general ILP problem. We then show simulations of these machines using MATLAB running on a single core of a Xeon processor for several ILP benchmark problems taken from the MIPLIB 2010 library, and compare our results against a renowned commercial solver. In particular, we find within minutes feasible solutions for one of these hard problems (f2000 from MIPLIB 2010) whose feasibility, to the best of our knowledge, has remained unknown for the past eight years.
Link: https://arxiv.org/abs/1808.09999
====================================================
An Operation Sequence Model for Explainable Neural Machine Translation (Felix Stahlberg - 29 August, 2018)
Our technique can outperform a plain text system in terms of BLEU score under the recent Transformer architecture on Japanese-English and Portuguese-English, and is within 0.5 BLEU difference on Spanish-English.
Link: https://arxiv.org/abs/1808.09688
====================================================
Towards Semi-Supervised Learning for Deep Semantic Role Labeling (Sanket Vaibhav Mehta - 28 August, 2018)
On CoNLL-2012 English section, the proposed semi-supervised training with 1%, 10% SRL-labeled data and varying amounts of SRL-unlabeled data achieves +1.58, +0.78 F1, respectively, over the pre-trained models that were trained on SOTA architecture with ELMo on the same SRL-labeled data. Additionally, by using the syntactic-inconsistency loss on inference time, the proposed model achieves +3.67, +2.1 F1 over pre-trained model on 1%, 10% SRL-labeled data, respectively.
Link: https://arxiv.org/abs/1808.09543
====================================================
Internet of Things: Technology, Applications and Standardardization (Jaydip Sen - 25 August, 2018)
This book presents some of the state-of-the-art research work in the field of the IoT, especially on the issues of communication protocols, interoperability of protocols and semantics, trust security and privacy issues, reference architecture design, and standardization
Link: https://arxiv.org/abs/1808.09390
====================================================
Bridging Knowledge Gaps in Neural Entailment via Symbolic Models (Dongyeop Kang - 4 September, 2018)
Our new architecture combines standard neural entailment models with a knowledge lookup module. On the SciTail dataset, NSnet outperforms a simpler combination of the two predictions by 3% and the base entailment model by 5%.
Link: https://arxiv.org/abs/1808.09333
====================================================
Joint Aspect and Polarity Classification for Aspect-based Sentiment Analysis with End-to-End Neural Networks (Martin Schmitt - 28 August, 2018)
We conduct experiments with different neural architectures and word representations on the recent GermEval 2017 dataset. The combination of a convolutional neural network and fasttext embeddings outperformed the best submission of the shared task in 2017, establishing a new state of the art.
Link: https://arxiv.org/abs/1808.09238
====================================================
Guided Neural Language Generation for Abstractive Summarization using Abstract Meaning Representation ( Hardy - 28 August, 2018)
Recent work on abstractive summarization has made progress with neural encoder-decoder architectures. We demonstrate that this guidance improves summarization results by 7.4 and 10.5 points in ROUGE-2 using gold standard AMR parses and parses obtained from an off-the-shelf parser respectively. We also find that the summarization performance using the latter is 2 ROUGE-2 points higher than that of a well-established neural encoder-decoder approach trained on a larger dataset
Link: https://arxiv.org/abs/1808.09160
====================================================
Single Shot Scene Text Retrieval (Lluís Gómez - 27 August, 2018)
Our experiments demonstrate that the proposed architecture outperforms previous state-of-the-art while it offers a significant increase in processing speed.
Link: https://arxiv.org/abs/1808.09044
====================================================
Pyramidal Recurrent Unit for Language Modeling (Sachin Mehta - 27 August, 2018)
This architecture gives strong results on word-level language modeling while reducing the number of parameters significantly. (2018) by up to 1.3 points while learning 15-20% fewer parameters
Link: https://arxiv.org/abs/1808.09029
====================================================
Review Helpfulness Assessment based on Convolutional Neural Network (Xianshan Qu - 27 August, 2018)
To our knowledge, this is the first use of this architecture to address this problem. We demonstrate that this can improve the overall accuracy by 2%. Finally, we evaluate the method on a benchmark dataset and show an improvement in accuracy relative to published results for traditional methods of 2.5% for a model trained using only review text and 4.24% for a model trained on a combination of rating star information and review text.
Link: https://arxiv.org/abs/1808.09016
====================================================
Why Self-Attention? A Targeted Evaluation of Neural Machine Translation Architectures (Gongbo Tang - 28 August, 2018)
Recently, non-recurrent architectures (convolutional, self-attentional) have outperformed RNNs in neural machine translation. Our experimental results show that: 1) self-attentional networks and CNNs do not outperform RNNs in modeling subject-verb agreement over long distances; 2) self-attentional networks perform distinctly better than RNNs and CNNs on word sense disambiguation.
Link: https://arxiv.org/abs/1808.08946
====================================================
Deep Learning for Stress Field Prediction Using Convolutional Neural Networks (Zhenguo Nie - 27 August, 2018)
Both the tow architectures are stable and converged reliably in training and testing on GPUs. Mean relative error (MRE) of the SE-Res-FCN model is about 0.25% with respect to the average ground truth
Link: https://arxiv.org/abs/1808.08914
====================================================
Attentive Sequence to Sequence Translation for Localizing Clips of Interest by Natural Language Descriptions (Ke Ning - 27 August, 2018)
The hierarchical architecture exploits video content with multiple granularities, ranging from subtle details to global context. Our ASST outperforms the state-of-the-art by $4.28\%$ in Rank$@1$ on the DiDeMo dataset. On the Charades-STA dataset, we significantly improve the state-of-the-art by $13.41\%$ in Rank$@1,IoU=0.5$.
Link: https://arxiv.org/abs/1808.08803
====================================================
An Empirical Study of Architecting for Continuous Delivery and Deployment (Mojtaba Shahin - 27 August, 2018)
Whilst an increasing amount of the literature covers different aspects of CD, little is known about the role of software architecture in CD and how an application should be (re-) architected to enable and support CD. We have conducted a mixed-methods empirical study that collected data through in-depth, semi-structured interviews with 21 industrial practitioners from 19 organizations, and a survey of 91 professional software practitioners
Link: https://arxiv.org/abs/1808.08796
====================================================
Adaptive Structural Learning of Deep Belief Network for Medical Examination Data and Its Knowledge Extraction by using C4.5 (Shin Kamada - 27 August, 2018)
Deep Learning has a hierarchical network architecture to represent the complicated feature of input patterns. The prediction system shows higher classification accuracy (99.8% for training and 95.5% for test) than the traditional DBN
Link: https://arxiv.org/abs/1808.08777
====================================================
Don't Give Me the Details, Just the Summary! Topic-Aware Convolutional Neural Networks for Extreme Summarization (Shashi Narayan - 27 August, 2018)
We demonstrate experimentally that this architecture captures long-range dependencies in a document and recognizes pertinent content, outperforming an oracle extractive system and state-of-the-art abstractive approaches when evaluated automatically and by humans.
Link: https://arxiv.org/abs/1808.08745
====================================================
Automatic 3D bi-ventricular segmentation of cardiac images by a shape-constrained multi-task deep learning approach (Jinming Duan - 28 August, 2018)
The architecture of the proposed FCN uses a 2.5D representation, thus combining the computational advantage of 2D FCNs networks and the capability of addressing 3D spatial consistency without compromising segmentation accuracy. We validate the pipeline on 1831 healthy subjects and 649 subjects with pulmonary hypertension
Link: https://arxiv.org/abs/1808.08578
====================================================
Rain Streak Removal for Single Image via Kernel Guided CNN (Ye-Tao Wang - 28 August, 2018)
In this paper, we propose a novel rain streak removal approach using a kernel guided convolutional neural network (KGCNN), achieving the state-of-the-art performance with simple network architectures
Link: https://arxiv.org/abs/1808.08545
====================================================
Event Detection with Neural Networks: A Rigorous Empirical Evaluation (J. Walker Orr - 26 August, 2018)
While the neural network models have generally led the state-of-the-art, the differences in performance between different architectures have not been rigorously studied
Link: https://arxiv.org/abs/1808.08504
====================================================
Painting Outside the Box: Image Outpainting with GANs (Mark Sabini - 25 August, 2018)
We use a three-phase training schedule to stably train a DCGAN architecture on a subset of the Places365 dataset. Once trained, our model is able to outpaint $128 \times 128$ color images relatively realistically, thus allowing for recursive outpainting
Link: https://arxiv.org/abs/1808.08483
====================================================
Privacy in Internet of Things: from Principles to Technologies (Chao Li - 25 August, 2018)
In this paper, we review the state-of-the-art principles of privacy laws, the architectures for IoT and the representative privacy enhancing technologies (PETs)
Link: https://arxiv.org/abs/1808.08443
====================================================
Deep Convolutional Neural Network with Mixup for Environmental Sound Classification (Zhichao Zhang - 25 August, 2018)
Our network architecture uses stacked convolutional and pooling layers to extract high-level feature representations from spectrogram-like features. Our experimental results demonstrated that our ESC system has achieved the state-of-the-art performance (83.7%) on UrbanSound8K and competitive performance on ESC-50 and ESC-10.
Link: https://arxiv.org/abs/1808.08405
====================================================
A Novel Deep Neural Network Architecture for Mars Visual Navigation (Jiang Zhang - 25 August, 2018)
By employing this architecture, Mars rover can determine the optimal navigation policy to the target point directly from original Martian environment images. Moreover, compared with the existing state-of-the-art algorithm, the training time is reduced by 45.8%
Link: https://arxiv.org/abs/1808.08395
====================================================
Improving Breast Cancer Detection using Symmetry Information with Deep Learning (Yeman Brhane Hagos - 17 August, 2018)
At candidate level, AUC value of 0.933 with 95% confidence interval of [0.920, 0.954] was obtained when symmetry information is incorporated in comparison with baseline architecture which yielded AUC value of 0.929 with [0.919, 0.947] confidence interval. By incorporating symmetrical information, although there was no a significant candidate level performance again (p = 0.111), we have found a compelling result at exam level with CPM value of 0.733 (p = 0.001)
Link: https://arxiv.org/abs/1808.08273
====================================================
Recalibrating Fully Convolutional Networks with Spatial and Channel 'Squeeze & Excitation' Blocks (Abhijit Guha Roy - 23 August, 2018)
The recalibration is achieved by simple computational blocks that can be easily integrated in F-CNNs architectures. Importantly, SE blocks only lead to a minimal increase in model complexity of about 1.5%, while the Dice score increases by 4-9% in the case of U-Net
Link: https://arxiv.org/abs/1808.08127
====================================================
Left ventricle quantification through spatio-temporal CNNs (Alejandro Debus - 23 August, 2018)
We show that incorporating such information by means of spatio-temporal convolutions into standard LV quantification architectures improves the accuracy of the predictions when compared with single-slice models, achieving competitive results for all cardiac indices and significantly breaking the state of the art (Xue et al., 2018, MedIA) for cardiac phase estimation.
Link: https://arxiv.org/abs/1808.07967
====================================================
Financial Aspect-Based Sentiment Analysis using Deep Representations (Steve Yang - 23 August, 2018)
FiQA contains high quality labels, but it still lacks data quantity to apply traditional ABSA deep learning architecture. Our results show an 8.7% improvement in the F1 score for classification and an 11% improvement over the MSE for regression on current state-of-the-art results.
Link: https://arxiv.org/abs/1808.07931
====================================================
Progressive Deep Neural Networks Acceleration via Soft Filter Pruning (Yang He - 23 August, 2018)
Moreover, our approach has been demonstrated effective for many advanced CNN architectures. Notably, on ILSCRC-2012, our method reduces more than 42% FLOPs on ResNet-101 with even 0.2% top-5 accuracy improvement, which has advanced the state-of-the-art. On ResNet-50, our progressive pruning method have 1.08% top-1 accuracy improvement over the pruning method without progressive pruning.
Link: https://arxiv.org/abs/1808.07471
====================================================
3D Topology Optimization using Convolutional Neural Networks (Saurabh Banga - 22 August, 2018)
To this end, we propose a deep learning approach based on a 3D encoder-decoder Convolutional Neural Network architecture for accelerating 3D topology optimization and to determine the optimal computational strategy for its deployment. For the best performing network, we achieved about 40% reduction in overall computation time while also attaining structural accuracies in the order of 96%.
Link: https://arxiv.org/abs/1808.07440
====================================================
DeepCorr: Strong Flow Correlation Attacks on Tor Using Deep Learning (Milad Nasr - 22 August, 2018)
DeepCorr leverages an advanced deep learning architecture to learn a flow correlation function tailored to Tor's complex network this is in contrast to previous works' use of generic statistical correlation metrics to correlated Tor flows. For instance, by collecting only about 900 packets of each target Tor flow (roughly 900KB of Tor data), DeepCorr provides a flow correlation accuracy of 96% compared to 4% by the state-of-the-art system of RAPTOR using the same exact setting.
Link: https://arxiv.org/abs/1808.07285
====================================================
Neural Architecture Optimization (Renqian Luo - 5 September, 2018)
Experiments show that the architecture discovered by our method is very competitive for image classification task on CIFAR-10 and language modeling task on PTB, outperforming or on par with the best results of previous architecture search methods with a significantly reduction of computational resources. Specifically we obtain $2.07\%$ test set error rate for CIFAR-10 image classification task and $55.9$ test set perplexity of PTB language modeling task
Link: https://arxiv.org/abs/1808.07233
====================================================
On Deep Neural Networks for Detecting Heart Disease (Nathalie-Sofia Tomov - 21 August, 2018)
The HEARO-5 architecture, yielding 99% accuracy and 0.98 MCC, significantly outperforms currently published research in the area.
Link: https://arxiv.org/abs/1808.07168
====================================================
Language Identification in Code-Mixed Data using Multichannel Neural Networks and Context Capture (Soumil Mandal - 21 August, 2018)
Inspired from the recent advancements in neural network architectures for computer vision tasks, we have implemented multichannel neural networks combining CNN and LSTM for word level language identification of code-mixed data. Combining this with a Bi-LSTM-CRF context capture module, accuracies of 93.28% and 93.32% is achieved on our two testing sets.
Link: https://arxiv.org/abs/1808.07118
====================================================
QuAC : Question Answering in Context (Eunsol Choi - 27 August, 2018)
We also report results for a number of reference models, including a recently state-of-the-art reading comprehension architecture extended to model dialog context. Our best model underperforms humans by 20 F1, suggesting that there is significant room for future work on this data
Link: https://arxiv.org/abs/1808.07036
====================================================
Soft Filter Pruning for Accelerating Deep Convolutional Neural Networks (Yang He - 21 August, 2018)
Moreover, our approach has been demonstrated effective for many advanced CNN architectures. Notably, on ILSCRC-2012, SFP reduces more than 42% FLOPs on ResNet-101 with even 0.2% top-5 accuracy improvement, which has advanced the state-of-the-art
Link: https://arxiv.org/abs/1808.06866
====================================================
Deep Learned Full-3D Object Completion from Single View (Dario Rethage - 21 August, 2018)
This writing proposes a new approach to 3D reconstruction and scene understanding, which implicitly learns 3D geometry from depth maps pairing a deep convolutional neural network architecture with an auto-encoder. The relatively small network, consisting of roughly 4 million weights, achieves a 92.9% reconstruction accuracy at a 30x30x30 resolution through the use of a pre-trained decompression layer
Link: https://arxiv.org/abs/1808.06843
====================================================
Real-time Analog Pixel-to-pixel Dynamic Frame Differencing with Memristive Sensing Circuits (Olga Krestinskaya - 21 August, 2018)
The proposed circuit is integrated into a pixel-parallel and pixel-column architectures. The power dissipation of the proposed circuit is $96.64mW$, and on-chip ares is $531.66 μm^2$
Link: https://arxiv.org/abs/1808.06780
====================================================
Constrained-size Tensorflow Models for YouTube-8M Video Understanding Challenge (Tianqi Liu - 12 September, 2018)
For each single model, we use the same network architecture as in the winning solution of the first YouTube-8M video understanding competition, namely Gated NetVLAD. We train the single models separately in tensorflow's default float32 precision, then replace weights with float16 precision and ensemble them in the evaluation and inference stages., achieving 48.5% compression rate without loss of precision. Our best model achieved 88.324% GAP on private leaderboard
Link: https://arxiv.org/abs/1808.06739
====================================================
Fast Spectrogram Inversion using Multi-head Convolutional Neural Networks (Sercan O. Arik - 20 August, 2018)
We propose the multi-head convolutional neural network (MCNN) architecture for waveform synthesis from spectrograms. MCNN achieves more than an order of magnitude higher compute intensity than commonly-used iterative algorithms like Griffin-Lim, yielding efficient utilization for modern multi-core processors, and very fast (more than 300x real-time) waveform synthesis
Link: https://arxiv.org/abs/1808.06719
====================================================
VERAM: View-Enhanced Recurrent Attention Model for 3D Shape Classification (Songle Chen - 20 August, 2018)
Taking grayscale image as input and AlexNet as CNN architecture, VERAM with 9 views achieves instance-level and class-level accuracy of 95:5% and 95:3% on ModelNet10, 93:7% and 92:1% on ModelNet40, both are the state-of-the-art performance under the same number of views.
Link: https://arxiv.org/abs/1808.06698
====================================================
Class2Str: End to End Latent Hierarchy Learning (Soham Saha - 20 August, 2018)
We show that for some of the best performing architectures on CIFAR and Imagenet datasets, the proposed replacement and training by LH classifier recovers the accuracy, with a fraction of the number of parameters in the classifier part. Compared to the previous work of HDCNN, which also learns a 2 level hierarchy, we are able to learn a hierarchy at an arbitrary number of levels as well as obtain an accuracy improvement on the Imagenet classification task over them
Link: https://arxiv.org/abs/1808.06675
====================================================
A Hybrid Differential Evolution Approach to Designing Deep Convolutional Neural Networks for Image Classification (Bin Wang - 21 August, 2018)
Firstly, an existing effective encoding scheme is refined to cater for variable-length CNN architectures; Secondly, the new mutation and crossover operators are developed for variable-length DE to optimise the hyperparameters of CNNs; Finally, the new second crossover is introduced to evolve the depth of the CNN architectures. The proposed algorithm is tested on six widely-used benchmark datasets and the results are compared to 12 state-of-the-art methods, which shows the proposed method is vigorously competitive to the state-of-the-art algorithms
Link: https://arxiv.org/abs/1808.06661
====================================================
Video-to-Video Synthesis (Ting-Chun Wang - 20 August, 2018)
Through carefully-designed generator and discriminator architectures, coupled with a spatio-temporal adversarial objective, we achieve high-resolution, photorealistic, temporally coherent video results on a diverse set of input formats including segmentation masks, sketches, and poses. In particular, our model is capable of synthesizing 2K resolution videos of street scenes up to 30 seconds long, which significantly advances the state-of-the-art of video synthesis
Link: https://arxiv.org/abs/1808.06601
====================================================
CU-Net: Coupled U-Nets (Zhiqiang Tang - 20 August, 2018)
We design a new connectivity pattern for the U-Net architecture. However, it only has at least 60% fewer parameters than other approaches.
Link: https://arxiv.org/abs/1808.06521
====================================================
Deep Residual Network for Sound Source Localization in the Time Domain (Dmitry Suvorov - 20 August, 2018)
This study describes the gathered dataset and developed architecture of the neural network. The accuracy classification of 30 m sec sound frames is 99.2%. Its usage decreased word error rate by 1.14% in comparison with similar speech recognition pipeline using GCC-PHAT sound source localization.
Link: https://arxiv.org/abs/1808.06429
====================================================
Experimental Evaluation of Passive Optical Network Based Data Centre Architecture (AEA Eltraify - 18 August, 2018)
In this paper we report the implementation of a PON based data centre architecture that provides high resilience and high speed interconnections by providing alternative communication routes between servers in different racks. We implement the switching and routing functionalities within servers using 4x10GE Xilinx NetFPGA, and demonstrate end-to-end communication using IP cameras live video streaming over up to 100 km optical connections through WDM nodes and the PON network.
Link: https://arxiv.org/abs/1808.06143
====================================================
Energy Efficient Service Distribution in Internet of Things (Barzan Yosuf - 18 August, 2018)
We optimize the placement of the IoT services in this architecture so that the total power consumption is minimized. Our results show that, introducing local computation at the IoT layer can bring up to 90% power savings compared with general purpose servers in a central cloud.
Link: https://arxiv.org/abs/1808.06120
====================================================
Compiler Enhanced Scheduling for OpenMP for Heterogeneous Multiprocessors (Jyothi Krishna V S - 18 August, 2018)
We implement a compiler for OpenMP and demonstrate its efficiency in Samsung Exynos with big.LITTLE architecture. On an average, we see 18% reduction in runtime and 14% reduction in energy consumption in standard NPB and FSU benchmarks with CES across multiple frequencies and core configurations in big.LITTLE.
Link: https://arxiv.org/abs/1808.06074
====================================================
CellLineNet: End-to-End Learning and Transfer Learning For Multiclass Epithelial Breast cell Line Classification via a Convolutional Neural Network (Darlington Ahiale Akogo - 18 August, 2018)
Using Transfer Learning, the 28-layer MobileNet Convolutional Neural Network architecture with pre-trained ImageNet weights is extended and fine tuned to the Multiclass Epithelial Breast cell Line Classification problem. CellLineNet simply requires an imaged Cell Line as input and it outputs the type of breast epithelial cell line (MDA-MB-468, MCF7, 10A, 12A or HC11) as predicted probabilities for the 5 classes. CellLineNet scored a 96.67% Accuracy.
Link: https://arxiv.org/abs/1808.06041
====================================================
Neuromorphic Architecture for the Hierarchical Temporal Memory (Abdullah M. Zyarah - 17 August, 2018)
The spatial pooler architecture is synthesized on Xilinx ZYNQ-7, with 91.16% classification accuracy for MNIST and 90\% accuracy for EUNF, with noise. For the temporal memory sequence prediction, first and second order predictions are observed for a 5-number long sequence generated from EUNF dataset and 95% accuracy is obtained
Link: https://arxiv.org/abs/1808.05839
====================================================
Read + Verify: Machine Reading Comprehension with Unanswerable Questions (Minghao Hu - 5 September, 2018)
Moreover, we introduce two auxiliary losses to help the reader better handle answer extraction as well as no-answer detection, and investigate three different architectures for the answer verifier. Our experiments on the SQuAD 2.0 dataset show that our system achieves a score of 74.2 F1 on the test set, outperforming all previous approaches at the time of submission (Aug
Link: https://arxiv.org/abs/1808.05759
====================================================
Dynamic Routing on Deep Neural Network for Thoracic Disease Classification and Sensitive Area Localization (Yan Shen - 17 August, 2018)
We address this problem by utilizing the recently developed routing-by agreement mechanism in our architecture. We demonstrate our results on the NIH chestX-ray14 dataset that consists of 112,120 images on 30,805 unique patients including 14 kinds of lung diseases.
Link: https://arxiv.org/abs/1808.05744
====================================================
BlockQNN: Efficient Block-wise Neural Network Architecture Generation (Zhao Zhong - 16 August, 2018)
However, most usable network architectures are hand-crafted and usually require expertise and elaborate design. The block-wise generation brings unique advantages: (1) it yields state-of-the-art results in comparison to the hand-crafted networks on image classification, particularly, the best network generated by BlockQNN achieves 2.35% top-1 error rate on CIFAR-10. (2) it offers tremendous reduction of the search space in designing networks, spending only 3 days with 32 GPUs. A faster version can yield a comparable result with only 1 GPU in 20 hours. The best network achieves very competitive accuracy of 82.0% top-1 and 96.0% top-5 on ImageNet.
Link: https://arxiv.org/abs/1808.05584
====================================================
LARNN: Linear Attention Recurrent Neural Network (Guillaume Chevalier - 16 August, 2018)
This neural architecture yields better results than the vanilla LSTM cells. It can obtain results of 91.92% for the test accuracy, compared to the previously attained 91.65% using vanilla LSTM cells. Note that this is not to compare to other research, where up to 93.35% is obtained, but costly using 18 LSTM cells rather than with 2 to 3 cells as analyzed here
Link: https://arxiv.org/abs/1808.05578
====================================================
Deeper Image Quality Transfer: Training Low-Memory Neural Networks for 3D Images (Stefano B. Blumberg - 16 August, 2018)
We exploit memory-efficient backpropagation techniques, to reduce the memory complexity of network training from being linear in the network's depth, to being roughly constant $ - $ permitting us to elongate deep architectures with negligible memory increase. We obtain substantially better results than the previous state-of-the-art model with a slight memory increase, reducing the root-mean-squared-error by $ 13\% $
Link: https://arxiv.org/abs/1808.05577
====================================================
Anatomy Of High-Performance Deep Learning Convolutions On SIMD Architectures (Evangelos Georganas - 20 August, 2018)
In this paper, we introduce direct convolution kernels for x86 architectures, in particular for Xeon and XeonPhi systems, which are implemented via a dynamic compilation approach
Link: https://arxiv.org/abs/1808.05567
====================================================
DNN Feature Map Compression using Learned Representation over GF(2) (Denis A. Gudovskiy - 15 August, 2018)
We apply the proposed network architectures derived from modified SqueezeNet and MobileNetV2 to the tasks of ImageNet classification and PASCAL VOC object detection. Compared to prior approaches, the conducted experiments show a factor of 2 decrease in memory requirements with minor degradation in accuracy while adding only bitwise computations.
Link: https://arxiv.org/abs/1808.05285
====================================================
Using Regular Languages to Explore the Representational Capacity of Recurrent Neural Architectures (Abhijit Mahalunkar - 15 August, 2018)
In order to test these state-of-the-art architectures, there is growing need for rich benchmarking datasets
Link: https://arxiv.org/abs/1808.05128
====================================================
Ensemble of Convolutional Neural Networks for Dermoscopic Images Classification (Tomáš Majtner - 15 August, 2018)
The proposed solution is based on deep learning, where we employed transfer learning strategy on VGG16 and GoogLeNet architectures. The solution was evaluated on Task 3: Lesion Diagnosis of the ISIC 2018: Skin Lesion Analysis Towards Melanoma Detection.
Link: https://arxiv.org/abs/1808.05071
====================================================
Deep EHR: Chronic Disease Prediction Using Medical Notes (Jingshu Liu - 14 August, 2018)
We compareperformance of different deep learning architectures including CNN, LSTM and hierarchical models.In contrast to traditional text-based prediction models, our approach does not require disease specificfeature engineering, and can handle negations and numerical values that exist in the text. Ourresults on a cohort of about 1 million patients show that models using text outperform modelsusing just structured data, and that models capable of using numerical values and negations in thetext, in addition to the raw text, further improve performance
Link: https://arxiv.org/abs/1808.04928
====================================================
Microservices in Practice: A Survey Study (Markos Viggiato - 14 August, 2018)
Microservices architectures have become largely popular in the last years. Thereupon, in this paper, we report the results of a survey with 122 professionals who work with microservices
Link: https://arxiv.org/abs/1808.04836
====================================================
DeepNeuro: an open-source deep learning toolbox for neuroimaging (Andrew Beers - 14 August, 2018)
We show how this framework can be used to both design and train neural network architectures, as well as modify state-of-the-art architectures in a flexible and intuitive way
Link: https://arxiv.org/abs/1808.04589
====================================================
Automatic Airway Segmentation in chest CT using Convolutional Neural Networks (A. Garcia-Uceda Juarez - 14 August, 2018)
Recently, deep convolutional neural networks (CNNs) have become the state-of-the-art for many segmentation tasks, and in particular the so-called Unet architecture for biomedical images. The method is trained on a dataset composed of 12 CTs, and tested on another 6 CTs. We evaluate the influence of different loss functions and data augmentation techniques, and reach an average dice coefficient of 0.8 between the ground-truth and our automated segmentations.
Link: https://arxiv.org/abs/1808.04576
====================================================
Multimodal Deep Neural Networks using Both Engineered and Learned Representations for Biodegradability Prediction (Garrett B. Goh - 13 September, 2018)
In this work, we develop a novel multimodal CNN-MLP neural network architecture that utilizes both domain-specific feature engineering as well as learned representations from raw data. DeepBioD, a multimodal CNN-MLP network is more accurate than either standalone network designs, and achieves an error classification rate of 0.125 that is 27% lower than the current state-of-the-art
Link: https://arxiv.org/abs/1808.04456
====================================================
Murmur Detection Using Parallel Recurrent & Convolutional Neural Networks (Shahnawaz Alam - 13 August, 2018)
We introduce a novel deep neural network architecture using parallel combination of the Recurrent Neural Network (RNN) based Bidirectional Long Short-Term Memory (BiLSTM) & Convolutional Neural Network (CNN) to learn visual and time-dependent characteristics of Murmur in PCG waveform. The proposed method was evaluated on a large dataset using 5-fold cross-validation, resulting in a sensitivity and specificity of 96 +- 0.6 % , 100 +- 0 % respectively and F1 Score of 98 +- 0.3 %.
Link: https://arxiv.org/abs/1808.04411
====================================================
A Domain Guided CNN Architecture for Predicting Age from Structural Brain Images (Pascal Sturmfels - 11 August, 2018)
We present two simple modifications to existing CNN architectures based on brain image structure. Applied to the task of brain age prediction, our network achieves a mean absolute error (MAE) of 1.4 years and trains 30% faster than a CNN baseline that achieves a MAE of 1.6 years
Link: https://arxiv.org/abs/1808.04362
====================================================
iNNvestigate neural networks! (Maximilian Alber - 13 August, 2018)
To demonstrate the versatility of iNNvestigate, we provide an analysis of image classifications for variety of state-of-the-art neural network architectures.
Link: https://arxiv.org/abs/1808.04260
====================================================
A Reference Architecture for Datacenter Scheduling: Extended Technical Report (Georgios Andreadis - 13 August, 2018)
To demonstrate the validity of the reference architecture, we map to it state-of-the-art datacenter schedulers
Link: https://arxiv.org/abs/1808.04224
====================================================
Automatic Plaque Detection in IVOCT Pullbacks Using Convolutional Neural Networks (Nils Gessert - 24 August, 2018)
We fuse both representations in a multi-path architecture for more effective feature exploitation. Overall, we find that our combined model performs best with an accuracy of 91.7%, a sensitivity of 90.9% and a specificity of 92.4%
Link: https://arxiv.org/abs/1808.04187
====================================================
DenseRAN for Offline Handwritten Chinese Character Recognition (Wenchao Wang - 13 August, 2018)
In this study, we propose a novel radical analysis network with densely connected architecture (DenseRAN) to analyze Chinese character radicals and its two-dimensional structures simultaneously. Evaluated on ICDAR-2013 competition database, the proposed approach significantly outperforms whole-character modeling approach with a relative character error rate (CER) reduction of 18.54%. Meanwhile, for the case of recognizing 3277 unseen Chinese characters in CASIA-HWDB1.2 database, DenseRAN can achieve a character accuracy of about 41% while the traditional whole-character method has no capability to handle them.
Link: https://arxiv.org/abs/1808.04134
====================================================
Confidence penalty, annealing Gaussian noise and zoneout for biLSTM-CRF networks for named entity recognition (Antonio Jimeno Yepes - 13 August, 2018)
In this work, we have done an analysis of several methods that intend to optimize the performance of networks based on this architecture, which in some cases encourage overfitting avoidance. Results show that the optimization methods improve the performance of the biLSTM-CRF NER baseline system, setting a new state of the art performance for the CoNLL-2003 Spanish set with an F1 of 87.18.
Link: https://arxiv.org/abs/1808.04029
====================================================
Pervasive Attention: 2D Convolutional Neural Networks for Sequence-to-Sequence Prediction (Maha Elbayad - 11 August, 2018)
Current state-of-the-art machine translation systems are based on encoder-decoder architectures, that first encode the input sequence, and then generate an output sequence based on the input encoding
Link: https://arxiv.org/abs/1808.03867
====================================================
Self-Supervised Model Adaptation for Multimodal Semantic Segmentation (Abhinav Valada - 11 August, 2018)
In addition, we propose a computationally efficient unimodal segmentation architecture termed AdapNet++ that incorporates a new encoder with multiscale residual units and an efficient atrous spatial pyramid pooling that has a larger effective receptive field with more than 10x fewer parameters, complemented with a strong decoder with a multi-resolution supervision scheme that recovers high-resolution details. Comprehensive empirical evaluations on several benchmarks demonstrate that both our unimodal and multimodal architectures achieve state-of-the-art performance.
Link: https://arxiv.org/abs/1808.03833
====================================================
Automatically Designing CNN Architectures Using Genetic Algorithm for Image Classification (Yanan Sun - 11 August, 2018)
For some state-of-the-art CNNs, their architectures are hand-crafted with expertise in both CNNs and the investigated problems. Furthermore, the proposed algorithm also shows the competitive classification accuracy to the semi-automatic peer competitors, while reducing 10 times of the parameters
Link: https://arxiv.org/abs/1808.03818
====================================================
Dropout during inference as a model for neurological degeneration in an image captioning network (Bai Li - 10 August, 2018)
We replicate a variation of the image captioning architecture by Vinyals et al. We find that the generated sentences most closely approximate the word frequency distribution of the training corpus when using a moderate dropout of 0.4 during inference.
Link: https://arxiv.org/abs/1808.03747
====================================================
Error Forward-Propagation: Reusing Feedforward Connections to Propagate Errors in Deep Learning (Adam A. Kohan - 9 August, 2018)
This mechanism, Error Forward-Propagation, is a plausible basis for how error feedback occurs deep in the brain independent of and yet in support of the functionality underlying intricate network architectures. We show experimentally that recurrent neural networks with two and three hidden layers can be trained using Error Forward-Propagation on the MNIST and Fashion MNIST datasets, achieving $1.90\%$ and $11\%$ generalization errors respectively.
Link: https://arxiv.org/abs/1808.03357
====================================================
On-Chip Optical Convolutional Neural Networks (Hengameh Bagherian - 16 August, 2018)
In this paper, we propose a photonics circuit architecture which could consume a fraction of energy per inference compared with state of the art electronics.
Link: https://arxiv.org/abs/1808.03303
====================================================
Controllable Image-to-Video Translation: A Case Study on Facial Expression Generation (Lijie Fan - 8 August, 2018)
To this end, we design a novel neural network architecture that can incorporate the user input into its skip connections and propose several improvements to the adversarial training method for the neural network. Especially, we would like to highlight that even for the face images in the wild (downloaded from the Web and the authors' own photos), our model can generate high-quality facial expression videos of which about 50\% are labeled as real by Amazon Mechanical Turk workers.
Link: https://arxiv.org/abs/1808.02992
====================================================
SchiNet: Automatic Estimation of Symptoms of Schizophrenia from Facial Behaviour Analysis (Mina Bishay - 7 August, 2018)
We automatically analyse the facial behaviour of 91 out-patients - this is almost 3 times the number of patients in other studies - and propose SchiNet, a novel neural network architecture that estimates expression-related symptoms in two different assessment interviews
Link: https://arxiv.org/abs/1808.02531
====================================================
YOLO3D: End-to-end real-time 3D Oriented Object Bounding Box Detection from LiDAR Point Cloud (Waleed Ali - 7 August, 2018)
In this paper, we build on the success of the one-shot regression meta-architecture in the 2D perspective image space and extend it to generate oriented 3D object bounding boxes from LiDAR point cloud. Our results are showing promising figures on KITTI benchmark, achieving real-time performance (40 fps) on Titan X GPU.
Link: https://arxiv.org/abs/1808.02350
====================================================
Deep Learning for Domain Adaption: Engagement Recognition (Omid Mohamad Nezami - 7 August, 2018)
We find that our Transfer architecture outperforms standard deep learning architectures that we apply for the first time to engagement recognition, as well as approaches using HOG features and SVMs. The model achieves a classification accuracy of 72.38%, which is 6.1% better than the best baseline model on the test set of the ER dataset. Using the F1 measure and the area under the ROC curve, our Transfer model achieves 73.90% and 73.74%, exceeding the best baseline model by 3.49% and 5.33% respectively.
Link: https://arxiv.org/abs/1808.02324
====================================================
Fast Variance Reduction Method with Stochastic Batch Size (Xuanqing Liu - 6 August, 2018)
However, due to the presence of cache/disk IO effect in computer architecture, the number of data access cannot reflect the running time because of 1) random memory access is much slower than sequential access, 2) when data is too big to fit into memory, disk seeking takes even longer time. After taking these into account, choosing batch size of $1$ is no longer optimal, so we propose a new algorithm called SAGA++ and show how to calculate the optimal average batch size theoretically
Link: https://arxiv.org/abs/1808.02169
====================================================
Efficient Fusion of Sparse and Complementary Convolutions (Chun-Fu Chen - 10 September, 2018)
We integrate this module into various network architectures and demonstrate its effectiveness on three vision tasks, object classification, localization and detection. For object detection, our approach leads to a VGG-16-based Faster RCNN detector that is 12.4$\times$ smaller and about 3$\times$ faster than the baseline.
Link: https://arxiv.org/abs/1808.02167
====================================================
Residual Memory Networks: Feed-forward approach to learn long temporal dependencies (Murali Karthick Baskar - 6 August, 2018)
In this paper we propose a residual memory neural network (RMN) architecture to model short-time dependencies using deep feed-forward layers having residual and time delayed connections. Recognition performance of RMN trained with 300 hours of Switchboard corpus is compared with various state-of-the-art LVCSR systems. The results indicate that RMN and BRMN gains 6 % and 3.8 % relative improvement over LSTM and BLSTM networks.
Link: https://arxiv.org/abs/1808.01916
====================================================
Classification of Dermoscopy Images using Deep Learning (Nithin D Reddy - 5 August, 2018)
We trained a convolutional neural network based on the ResNet50 architecture to accurately classify dermoscopy images of skin lesions into one of seven disease categories. Using our custom model, we obtained a balanced accuracy of 91% on the validation dataset.
Link: https://arxiv.org/abs/1808.01607
====================================================
A Deep Learning based Joint Segmentation and Classification Framework for Glaucoma Assesment in Retinal Color Fundus Images (Arunava Chakravarty - 29 July, 2018)
The use of fewer network parameters and the sharing of the CNN features for multiple related tasks ensures the good generalizability of the architecture, allowing it to be trained on small training sets. The cross-testing performance of the proposed method on an independent validation set acquired using a different camera and image resolution was found to be good with an average dice score of 0.92 for OD, 0.84 for OC and AUC of 0.95 on the task of glaucoma classification illustrating its potential as a mass screening tool for the early detection of glaucoma.
Link: https://arxiv.org/abs/1808.01355
====================================================
SWDE : A Sub-Word And Document Embedding Based Engine for Clickbait Detection (Vaibhav Kumar - 2 August, 2018)
We generate sub-word level embeddings of the title using Convolutional Neural Networks and use them to train a bidirectional LSTM architecture. We test our model over 2538 posts (having trained it on 17000 records) and achieve an accuracy of 83.49% outscoring previous state-of-the-art approaches.
Link: https://arxiv.org/abs/1808.00957
====================================================
BiSeNet: Bilateral Segmentation Network for Real-time Semantic Segmentation (Changqian Yu - 2 August, 2018)
The proposed architecture makes a right balance between the speed and segmentation performance on Cityscapes, CamVid, and COCO-Stuff datasets. Specifically, for a 2048x1024 input, we achieve 68.4% Mean IOU on the Cityscapes test dataset with speed of 105 FPS on one NVIDIA Titan XP card, which is significantly faster than the existing methods with comparable performance.
Link: https://arxiv.org/abs/1808.00897
====================================================
Weakly Supervised Localisation for Fetal Ultrasound Images (Nicolas Toussaint - 2 August, 2018)
We examine the use of convolutional neural network architectures coupled with soft proposal layers. Detection achieves an average accuracy of 90\% on individual regions, and show that the proposal maps correlate well with relevant anatomical structures
Link: https://arxiv.org/abs/1808.00793
====================================================
Approximate Probabilistic Neural Networks with Gated Threshold Logic (Olga Krestinskaya - 2 August, 2018)
The weights of the PNN are approximated using a memristive crossbar architecture. In particular, the proposed algorithm performs normalization of the training weights, and quantization into 16 levels which significantly reduces the complexity of the circuit.
Link: https://arxiv.org/abs/1808.00733
====================================================
Deep Learning for Radio Resource Allocation in Multi-Cell Networks (K. I. Ahmed - 2 August, 2018)
Starting with a brief overview of a deep neural network (DNN) as a DL model, relevant DNN architectures and the data training procedure, we provide an overview of existing state-of-the-art applying DL in the context of radio resource allocation. Simulation results show that the trained DL model is able to provide the desired optimal solution 86.3% of time.
Link: https://arxiv.org/abs/1808.00667
====================================================
The BaseJump Manycore Accelerator Network (Shaolin Xie - 15 August, 2018)
The BaseJump Manycore Accelerator-Network is an open source mesh-based On-Chip-Network which is designed leveraging the Bespoke Silicon Group's 20+ years of experience in designing manycore architectures. (2018), forming the basis of both a 1 GHz 496-core RISC-V manycore and a 10-core always-on low voltage complex
Link: https://arxiv.org/abs/1808.00650
====================================================
Efficient Progressive Neural Architecture Search (Juan-Manuel Perez-Rua - 1 August, 2018)
We propose a method that aggregates two main results of the previous state-of-the-art in neural architecture search. Sequential search has previously demonstrated its capabilities to find state-of-the-art neural architectures for image classification
Link: https://arxiv.org/abs/1808.00391
====================================================
Reinforced Evolutionary Neural Architecture Search (Yukang Chen - 6 September, 2018)
We conduct the proposed search method on CIFAR-10 with 4 GPUs (Titan Xp) across 1.5 days and discover a powerful network architecture
Link: https://arxiv.org/abs/1808.00193
====================================================
News Session-Based Recommendations using Deep Neural Networks (Gabriel de Souza P. Moreira - 16 September, 2018)
Experiments with an extensive number of session-based recommendation methods were performed and the proposed instantiation of CHAMELEON meta-architecture obtained a significant relative improvement in top-n accuracy and ranking metrics (10% on Hit Rate and 13% on MRR) over the best benchmark methods.
Link: https://arxiv.org/abs/1808.00076
====================================================
Modeling Task Effects in Human Reading with Neural Attention (Michael Hahn - 31 July, 2018)
We propose a neural architecture that combines an attention module (deciding whether to skip words) and a task module (memorizing the input). We show that our model predicts human skipping behavior, while also modeling reading times well, even though it skips 40% of the input
Link: https://arxiv.org/abs/1808.00054
====================================================
Compact Convolutional Neural Networks for Multi-Class, Personalised, Closed-Loop EEG-BCI (Pablo Ortega - 31 July, 2018)
Our preliminary results show that an efficient architecture (SmallNet), with only one convolutional layer, can classify 4 mental activities chosen by the user. It is kept up-to-date through the use of newly collected signals along playing, reaching an online accuracy of 47.6% where most approaches only report results obtained offline. Compared to our previous decoder of physiological signals relying on blinks, we increased by a factor 2 the amount of states among which the user can transit, bringing the opportunity for finer control of specific subtasks composing natural grasping in a self-paced way
Link: https://arxiv.org/abs/1807.11752
====================================================
Deep Cross Modal Learning for Caricature Verification and Identification(CaVINet) (Jatin Garg - 31 July, 2018)
This paper presents the first cross modal architecture that handles extreme distortions of caricatures using a deep learning network that learns similar representations across the modalities. The learned shared representation achieves 91% accuracy for verifying unseen images and 75% accuracy on unseen identities. Further, recognizing the identity in the image by knowledge transfer using a combination of shared and modality specific representations, resulted in an unprecedented performance of 85% rank-1 accuracy for caricatures and 95% rank-1 accuracy for visual images.
Link: https://arxiv.org/abs/1807.11688
====================================================
MnasNet: Platform-Aware Neural Architecture Search for Mobile (Mingxing Tan - 30 July, 2018)
In this paper, we propose an automated neural architecture search approach for designing resource-constrained mobile CNN models. On the ImageNet classification task, our model achieves 74.0% top-1 accuracy with 76ms latency on a Pixel phone, which is 1.5x faster than MobileNetV2 (Sandler et al. 2018) and 2.4x faster than NASNet (Zoph et al. 2018) with the same top-1 accuracy
Link: https://arxiv.org/abs/1807.11626
====================================================
Testing the Efficient Network TRaining (ENTR) Hypothesis: initially reducing training image size makes Convolutional Neural Network training for image recognition tasks more efficient (Thomas Cherico Wanger - 30 July, 2018)
Convolutional Neural Networks (CNN) for image recognition tasks are seeing rapid advances in the available architectures and how networks are trained based on large computational infrastructure and standard datasets with millions of images. We test this Efficient Network TRaining (ENTR) Hypothesis by training pre-trained Residual Network (ResNet) models (ResNet18, 34, & 50) on three small datasets (steel microstructures, bee images, and geographic aerial images) with a free cloud GPU
Link: https://arxiv.org/abs/1807.11583
====================================================
Harmonic-Percussive Source Separation with Deep Neural Networks and Phase Recovery (Konstantinos Drossos - 30 July, 2018)
MaD TwinNet is a deep learning architecture that has reached state-of-the-art results in monaural singing voice separation
Link: https://arxiv.org/abs/1807.11298
====================================================
Modular Sensor Fusion for Semantic Segmentation (Hermann Blum - 30 July, 2018)
Current multi-sensor deep learning based semantic segmentation approaches do not provide robustness to under-performing classes in one modality, or require a specific architecture with access to the full aligned multi-sensor training data. In our experiments, the approach improves performance in IoU over the best single modality segmentation results by up to 5%
Link: https://arxiv.org/abs/1807.11249
====================================================
Multi-Fiber Networks for Video Recognition (Yunpeng Chen - 18 September, 2018)
Extensive experimental results show that our multi-fiber architecture significantly boosts the efficiency of existing convolution networks for both image and video recognition tasks, achieving state-of-the-art performance on UCF-101, HMDB-51 and Kinetics datasets. Our proposed model requires over 9x and 13x less computations than the I3D and R(2+1)D models, respectively, yet providing higher accuracy.
Link: https://arxiv.org/abs/1807.11195
====================================================
Leveraging Medical Sentiment to Understand Patients Health on Social Media (Shweta Yadav - 30 July, 2018)
We propose an effective architecture that uses a Convolutional Neural Network (CNN) as a data-driven feature extractor and a Support Vector Machine (SVM) as a classifier. In addition to our dataset, we also evaluate our approach on the benchmark "CLEF eHealth 2014" corpora and show that our model outperforms the state-of-the-art techniques.
Link: https://arxiv.org/abs/1807.11172
====================================================
Towards Good Practices on Building Effective CNN Baseline Model for Person Re-identification (Fu Xiong - 29 July, 2018)
To answer this open question, we propose 3 good practices in this paper from the perspectives of adjusting CNN architecture and training procedure. The extensive experiments on 3 widely-used benchmark datasets demonstrate that, our propositions essentially facilitate the CNN baseline model to achieve the state-of-the-art performance without any other high-level domain knowledge or low-level technical trick.
Link: https://arxiv.org/abs/1807.11042
====================================================
Tiny-DSOD: Lightweight Object Detection for Resource-Restricted Usages (Yuxi Li - 29 July, 2018)
Tiny-DSOD introduces two innovative and ultra-efficient architecture blocks: depthwise dense block (DDB) based backbone and depthwise feature-pyramid-network (D-FPN) based front-end. We conduct extensive experiments on three famous benchmarks (PASCAL VOC 2007, KITTI, and COCO), and compare Tiny-DSOD to the state-of-the-art ultra-efficient object detection solutions such as Tiny-YOLO, MobileNet-SSD (v1 & v2), SqueezeDet, Pelee, etc. For instance, Tiny-DSOD achieves 72.1% mAP with only 0.95M parameters and 1.06B FLOPs, which is by far the state-of-the-arts result with such a low resource requirement.
Link: https://arxiv.org/abs/1807.11013
====================================================
PROPEL: Probabilistic Parametric Regression Loss for Convolutional Neural Networks (Muhammad Asad - 28 July, 2018)
PROPEL is fully differentiable and, hence, can be easily incorporated for end-to-end training of existing regressive CNN architectures. Our experimental results indicate that PROPEL significantly improves the performance of a CNN, while reducing model parameters by 10x as compared to the existing state-of-the-art.
Link: https://arxiv.org/abs/1807.10937
====================================================
Deep Leaf Segmentation Using Synthetic Data (Daniel Ward - 16 August, 2018)
We train a state-of-the-art deep learning segmentation architecture (Mask-RCNN) with a combination of real and synthetic images of Arabidopsis plants. Our proposed approach achieves 90% leaf segmentation score on the A1 test set outperforming the-state-of-the-art approaches for the CVPPP Leaf Segmentation Challenge (LSC). Our approach also achieves 81% mean performance over all five test datasets.
Link: https://arxiv.org/abs/1807.10931
====================================================
TBI Contusion Segmentation from MRI using Convolutional Neural Networks (Snehashis Roy - 27 July, 2018)
The CNN architecture proposed here was based on a state of the art CNN architecture from Google, called Inception. When compared with two recent TBI lesion segmentation methods, one based on CNN (called DeepMedic) and another based on random forests, the proposed algorithm showed improved segmentation accuracy on images of 18 patients with mild to severe TBI. Using a leave-one-out cross validation, the proposed model achieved a median Dice of 0.75, which was significantly better (p<0.01) than the two competing methods.
Link: https://arxiv.org/abs/1807.10839
====================================================
Improving Neural Sequence Labelling using Additional Linguistic Information (Mahtab Ahmed - 27 July, 2018)
Our proposed architecture achieves state of the art results on the benchmark datasets of POS, NER, and chunking
Link: https://arxiv.org/abs/1807.10805
====================================================
Network Compression using Correlation Analysis of Layer Responses (Xavier Suau - 28 September, 2018)
Both algorithms are evaluated against several architectures and datasets, and we show considerable compression rates without compromising accuracy, e.g., for VGG-16 on CIFAR-10 and CIFAR-100 PFA achieves a compression rate of 8x and 3x with an accuracy gain of 0.4% points and 1.4% points, respectively
Link: https://arxiv.org/abs/1807.10585
====================================================
Uncertainty and Interpretability in Convolutional Neural Networks for Semantic Segmentation of Colorectal Polyps (Kristoffer Wickstrøm - 16 July, 2018)
We evaluate and enhance several architectures of Fully Convolutional Networks (FCNs) for semantic segmentation of colorectal polyps and provide a comparison between these models. Our highest performing model achieves a 76.06\% mean IOU accuracy on the EndoScene dataset, a considerable improvement over the previous state-of-the-art.
Link: https://arxiv.org/abs/1807.10584
====================================================
Improving High Resolution Histology Image Classification with Deep Spatial Fusion Network (Yongxiang Huang - 27 July, 2018)
In this paper, we propose a novel architecture for automatic classification of high resolution histology images. The proposed method is evaluated using 10-fold cross-validation on 400 high resolution breast histology images with balanced labels and reports 95% accuracy on 4-class classification and 98.5% accuracy, 99.6% AUC on 2-class classification (carcinoma and non-carcinoma), which substantially outperforms previous methods and close to pathologist performance.
Link: https://arxiv.org/abs/1807.10552
====================================================
AXNet: ApproXimate computing using an end-to-end trainable neural network (Zhenghao Peng - 27 July, 2018)
Neural network based approximate computing is a universal architecture promising to gain tremendous energy-efficiency for many error resilient applications. Experiment results show 50.7% more invocation and substantial cuts of training time when compared to existing neural network based approximate computing framework.
Link: https://arxiv.org/abs/1807.10458
====================================================
DeepSPINE: Automated Lumbar Vertebral Segmentation, Disc-level Designation, and Spinal Stenosis Grading Using Deep Learning (Jen-Tang Lu - 26 July, 2018)
Specifically, we introduce three major contributions: (1) a natural-language-processing scheme to extract level-by-level ground-truth labels from free-text radiology reports for the various types and grades of spinal stenosis (2) accurate vertebral segmentation and disc-level localization using a U-Net architecture combined with a spine-curve fitting method, and (3) a multi-input, multi-task, and multi-class convolutional neural network to perform central canal and foraminal stenosis grading on both axial and sagittal imaging series inputs with the extracted report-derived labels applied to corresponding imaging level segments. This study uses a large dataset of 22796 disc-levels extracted from 4075 patients
Link: https://arxiv.org/abs/1807.10215
====================================================
UNet++: A Nested U-Net Architecture for Medical Image Segmentation (Zongwei Zhou - 18 July, 2018)
We have evaluated UNet++ in comparison with U-Net and wide U-Net architectures across multiple medical image segmentation tasks: nodule segmentation in the low-dose CT scans of chest, nuclei segmentation in the microscopy images, liver segmentation in abdominal CT scans, and polyp segmentation in colonoscopy videos. Our experiments demonstrate that UNet++ with deep supervision achieves an average IoU gain of 3.9 and 3.4 points over U-Net and wide U-Net, respectively.
Link: https://arxiv.org/abs/1807.10165
====================================================
MRI to FDG-PET: Cross-Modal Synthesis Using 3D U-Net For Multi-Modal Alzheimer's Classification (Apoorva Sikka - 30 July, 2018)
Current work focuses on a cross-modal approach to estimate FDG-PET scans for the given MR scans using a 3D U-Net architecture. The accuracy using only MRI is 70.18% while joint classification using synthesized PET and MRI is 74.43% with a p-value of $0.06$
Link: https://arxiv.org/abs/1807.10111
====================================================
Effects of Degradations on Deep Neural Network Architectures (Prasun Roy - 12 August, 2018)
The architecture has shown promising result by surpassing the performances of state-of-the-art CNN based models in some of the existing datasets. To the best of our knowledge, this is the first study on the performance of CapsuleNet (or CapsNet) and other state-of-the-art CNN architectures under different types of image degradations.
Link: https://arxiv.org/abs/1807.10108
====================================================
Towards a Deep Unified Framework for Nuclear Reactor Perturbation Analysis (Fabio De Sousa Ribeiro - 8 September, 2018)
512 dimensional representations were extracted from the 3D-CNN and LSTM architectures, and used as input to a fused multi-sigmoid classification layer to recognise the perturbation type
Link: https://arxiv.org/abs/1807.10096
====================================================
From Monolith to Microservices: A Classification of Refactoring Approaches (Jonas Fritzsch - 26 July, 2018)
One of the most discussed aspects in this context is finding the right service granularity to fully leverage the advantages of a Microservices architecture. This study first discusses the notion of architectural refactoring and subsequently compares 10 existing refactoring approaches recently proposed in academic literature
Link: https://arxiv.org/abs/1807.10059
====================================================
Multi-temporal Sentinel-1 and -2 Data Fusion for Optical Image Simulation (Wei He - 26 July, 2018)
The deep learning based methods that we chose to achieve the models are a convolutional neural network (CNN) with a residual architecture and a conditional generative adversarial network (cGAN). We validate our models using the Sentinel-1 and -2 datasets
Link: https://arxiv.org/abs/1807.09954
====================================================
Coreset-Based Neural Network Compression (Abhimanyu Dubey - 25 July, 2018)
Our method requires no retraining, is easy to implement, and obtains state-of-the-art compression performance across a wide variety of CNN architectures
Link: https://arxiv.org/abs/1807.09810
====================================================
End-to-End Incremental Learning (Francisco M. Castro - 3 September, 2018)
This is due to current neural network architectures requiring the entire dataset, consisting of all the samples from the old as well as the new classes, to update the model -a requirement that becomes easily unsustainable as the number of classes grows. We evaluate our method extensively on the CIFAR-100 and ImageNet (ILSVRC 2012) image classification datasets, and show state-of-the-art performance.
Link: https://arxiv.org/abs/1807.09536
====================================================
Toward Scale-Invariance and Position-Sensitive Region Proposal Networks (Hsueh-Fu Lu - 25 July, 2018)
The design of the network architecture aims to be simple while being effective and with real time performance. Without bells and whistles the proposed object proposal network significantly improves the AR at 1,000 proposals by $35\%$ and $45\%$ on PASCAL VOC and COCO dataset respectively and has a fast inference time of 44.8 ms for input image size of $640^{2}$
Link: https://arxiv.org/abs/1807.09528
====================================================
Millimeter-Wave Propagation within a Computer Chip Package (X. Timoneda - 25 July, 2018)
The WNoC paradigm has been extensively analyzed from the physical, network and architecture perspectives assuming mmWave band operation. Through parametric studies, a locally optimal configuration for 60 GHz WNoC is obtained, showing that chip-wide attenuation below 32.6 dB could be achieved with standard processes
Link: https://arxiv.org/abs/1807.09472
====================================================
Two at Once: Enhancing Learning and Generalization Capacities via IBN-Net (Xingang Pan - 27 July, 2018)
(2) IBN-Net can be applied to many advanced deep architectures, such as DenseNet, ResNet, ResNeXt, and SENet, and consistently improve their performance without increasing computational cost. With IBN-Net, we won the 1st place on the WAD 2018 Challenge Drivable Area track, with an mIoU of 86.18%.
Link: https://arxiv.org/abs/1807.09441
====================================================
Multicolumn Networks for Face Recognition (Weidi Xie - 24 July, 2018)
Comparing with the previous state-of-the-art architectures trained with the same dataset (VGGFace2), our Multicolumn Networks show an improvement of between 2-6% on the IARPA IJB face recognition benchmarks, and exceed the state of the art for all methods on these benchmarks.
Link: https://arxiv.org/abs/1807.09192
====================================================
StereoNet: Guided Hierarchical Refinement for Real-Time Edge-Aware Depth Prediction (Sameh Khamis - 23 July, 2018)
This paper presents StereoNet, the first end-to-end deep architecture for real-time stereo matching that runs at 60 fps on an NVidia Titan X, producing high-quality, edge-preserved, quantization-free disparity maps
Link: https://arxiv.org/abs/1807.08865
====================================================
CNN-based Facial Affect Analysis on Mobile Devices (Charlie Hewitt - 23 July, 2018)
Our results show that the proposed architectures retain similar performance to the dataset baseline while minimising storage requirements: achieving 58% accuracy for eight-class emotion classification and average RMSE of 0.39 for valence/arousal prediction
Link: https://arxiv.org/abs/1807.08775
====================================================
Automatic Speech Recognition for Humanitarian Applications in Somali (Raghav Menon - 23 July, 2018)
We evaluate several types of acoustic model, including recent neural architectures. We find that both types of data augmentation are beneficial to performance, with our best system using a combination of convolutional neural networks (CNNs), time-delay neural networks (TDNNs) and bi-directional long short term memory (BLSTMs) to achieve a word error rate of 53.75%.
Link: https://arxiv.org/abs/1807.08669
====================================================
Deep Learning from Label Proportions for Emphysema Quantification (Gerda Bortsova - 23 July, 2018)
Our architecture and loss combined improve the performance substantially (8% ICC) compared to a more conventional regression network. We outperform traditional lung densitometry and two recently published methods for emphysema quantification by a large margin (at least 7% AUC and 15% ICC), and achieve near-human-level performance
Link: https://arxiv.org/abs/1807.08601
====================================================
3D Convolutional Neural Networks for Tumor Segmentation using Long-range 2D Context (Pawel Mlynarski - 23 July, 2018)
Furthermore, we propose a network architecture in which the different MR sequences are processed by separate subnetworks in order to be more robust to the problem of missing MR sequences. We evaluate our method on the public benchmark of the BRATS 2017 challenge on the task of multiclass segmentation of malignant brain tumors. Our method achieves good performances and produces accurate segmentations with median Dice scores of 0.918 (whole tumor), 0.883 (tumor core) and 0.854 (enhancing core)
Link: https://arxiv.org/abs/1807.08599
====================================================
Hardware Optimizations of Dense Binary Hyperdimensional Computing: Rematerialization of Hypervectors, Binarized Bundling, and Combinational Associative Memory (Manuel Schmuck - 20 July, 2018)
The Pareto optimal HD architecture is mapped on only 18340 configurable logic blocks (CLBs) to learn and classify five hand gestures using four electromyography sensors.
Link: https://arxiv.org/abs/1807.08583
====================================================
Zero-shot keyword spotting for visual speech recognition in-the-wild (Themos Stafylakis - 25 July, 2018)
To this end, we devise an end-to-end architecture comprising (a) a state-of-the-art visual feature extractor based on spatiotemporal Residual Networks, (b) a grapheme-to-phoneme model based on sequence-to-sequence neural networks, and (c) a stack of recurrent neural networks which learn how to correlate visual features with the keyword representation
Link: https://arxiv.org/abs/1807.08469
====================================================
S3D: Single Shot multi-Span Detector via Fully 3D Convolutional Networks (Da Zhang - 7 August, 2018)
Our architecture, named S3D, encodes the entire video stream and discretizes the output space of temporal activity spans into a set of default spans over different temporal locations and scales. When evaluating on THUMOS'14 detection benchmark, S3D achieves state-of-the-art performance and is very efficient and can operate at 1271 FPS.
Link: https://arxiv.org/abs/1807.08069
====================================================
Surgical Phase Recognition of Short Video Shots Based on Temporal Modeling of Deep Features (Constantinos Loukas - 6 September, 2018)
We investigate four state-of-the-art CNN architectures (Alexnet, VGG19, GoogleNet, and ResNet101), for feature extraction via transfer learning. Most importantly, we investigate the role of 'elapsed time' (from the beginning of the operation), and we show that inclusion of this feature can increase performance dramatically (69% vs. 75% mean accuracy). Finally, a long short-term memory (LSTM) network was trained for video shot classification based on the fusion of CNN features with 'elapsed time', increasing the accuracy to 86%
Link: https://arxiv.org/abs/1807.07853
====================================================
Improving Image Clustering With Multiple Pretrained CNN Feature Extractors (Joris Guérin - 20 July, 2018)
We then propose a multi-input neural network architecture that is trained end-to-end to solve the MVC problem effectively. Our experimental results, conducted on three different natural image datasets, show that: 1. using multiple pretrained CNNs jointly as feature extractors improves image clustering; 2. using an end-to-end approach improves MVC; and 3
Link: https://arxiv.org/abs/1807.07760
====================================================
Fully Convolutional Pixel Adaptive Image Denoiser (Sungmin Cha - 19 July, 2018)
The three main contributions we make to significantly improve upon the original Neural AIDE are the followings; 1) implementing a novel fully convolutional architecture that boosts the base supervised model, 2) introducing data augmentation for adaptive fine-tuning to achieve much stronger adaptivity, and 3) proposing an effective unknown noise level estimation method
Link: https://arxiv.org/abs/1807.07569
====================================================
Attention-Guided Curriculum Learning for Weakly Supervised Classification and Localization of Thoracic Diseases on Chest Radiographs (Yuxing Tang - 19 July, 2018)
A two-path network architecture is designed to regress the heatmaps from selected seed samples in addition to the original classification task. AGCL achieves over 5.7\% (averaged over 14 diseases) increase in classification AUC and 7%/11% increases in Recall/Precision for the localization task compared to the state of the art.
Link: https://arxiv.org/abs/1807.07532
====================================================
Image Reconstruction via Variational Network for Real-Time Hand-Held Sound-Speed Imaging (Valery Vishnevskiy - 19 July, 2018)
For this problem, we herein present a variational network based image reconstruction architecture that is based on optimization loop unrolling, and provide an efficient training protocol of this network architecture on fully synthetic inclusion data. Complex inclusion geometries were shown to be successfully reconstructed, also improving over the prior-art by 23% in reconstruction error and by 10% in contrast on synthetic data. In a phantom study, we demonstrated the detection of multiple inclusions that were not distinguishable by prior-art reconstruction, meanwhile improving the contrast by 27% for a stiff inclusion and by 219% for a soft inclusion
Link: https://arxiv.org/abs/1807.07416
====================================================
Mr. DLib's Living Lab for Scholarly Recommendations (Joeran Beel - 19 July, 2018)
We present the architecture of Mr. During this time, 970,517 recommendations were delivered with a mean click-through rate of 0.22%.
Link: https://arxiv.org/abs/1807.07298
====================================================
Hierarchical Multi Task Learning With CTC (Ramon Sanabria - 25 July, 2018)
On the standard 300h Switchboard training setup, our hierarchical multi-task architecture exhibits improvements over single-task architectures with the same number of parameters. Our model obtains 14.0% Word Error Rate on the Eval2000 Switchboard subset without any decoder or language model, outperforming the current state-of-the-art on acoustic-to-word models.
Link: https://arxiv.org/abs/1807.07104
====================================================
Robot Learning in Homes: Improving Generalization and Reducing Dataset Bias (Abhinav Gupta - 18 July, 2018)
Our architecture which explicitly models the latent noise in the dataset also performed 10% better than one that did not factor out the noise
Link: https://arxiv.org/abs/1807.07049
====================================================
Crack-pot: Autonomous Road Crack and Pothole Detection (Sukhad Anand - 9 September, 2018)
The approach is based on a deep neural net architecture which detects cracks and potholes using texture and spatial features
Link: https://arxiv.org/abs/1810.05107
====================================================
Neural Relation Extraction Within and Across Sentence Boundaries (Pankaj Gupta - 11 October, 2018)
In this paper, we propose a novel architecture for this task: inter-sentential dependency-based neural networks (iDepNN)
Link: https://arxiv.org/abs/1810.05102
====================================================
Taming the Cross Entropy Loss (Manuel Martinez - 11 October, 2018)
We evaluate the TCE loss using the ResNet architecture on four image datasets that we artificially contaminated with various levels of label noise
Link: https://arxiv.org/abs/1810.05075
====================================================
USBCaptchaIn: Preventing (Un)Conventional Attacks from Promiscuously Used USB Devices in Industrial Control Systems (Federico Griscioli - 11 October, 2018)
The main component of the architecture we propose is an hardware, called USBCaptchaIn, intended to be in the middle between a critical machine and all USB devices
Link: https://arxiv.org/abs/1810.05005
====================================================
Location Dependency in Video Prediction (Niloofar Azizi - 11 October, 2018)
The effectiveness of location bias is evaluated on two architectures: Video Ladder Network (VLN) and Convolutional redictive Gating Pyramid (Conv-PGP)
Link: https://arxiv.org/abs/1810.04937
====================================================
Technical Report: KNN Joins Using a Hybrid Approach: Exploiting CPU/GPU Workload Characteristics (Michael Gowanlock - 10 October, 2018)
Since the CPU and GPU are considerably different architectures that are best exploited using different algorithms, we advocate for splitting the work between both architectures based on the characteristic workloads defined by the query points in the dataset. As such, we assign dense regions to the GPU, and sparse regions to the CPU to most efficiently exploit the relative strengths of each architecture
Link: https://arxiv.org/abs/1810.04758
====================================================
Training Generative Adversarial Networks with Binary Neurons by End-to-end Backpropagation (Hao-Wen Dong - 10 October, 2018)
We implement such a model to generate binarized MNIST digits and experimentally compare the performance for different types of binary neurons, GAN objectives and network architectures
Link: https://arxiv.org/abs/1810.04714
====================================================
uops.info: Characterizing Latency, Throughput, and Port Usage of Instructions on Intel Microarchitectures (Andreas Abel - 10 October, 2018)
Modern microarchitectures are some of the world's most complex man-made systems. As a consequence, it is increasingly difficult to predict, explain, let alone optimize the performance of software running on such microarchitectures
Link: https://arxiv.org/abs/1810.04610
====================================================
Time Efficient Data Migration among Clouds (Syeda Munazza Marium - 24 September, 2018)
Data migration researcher attempt to move data concerning various geographical locations, which contain huge data volumes, compact time limit and problematical architectures
Link: https://arxiv.org/abs/1810.04609
====================================================
Autonomous Mobility and Energy Service Management in Future Smart Cities: An Overview (Xiaoqi Tan - 7 October, 2018)
First, we propose a three-layer architecture for the convergence of future mobility and energy systems
Link: https://arxiv.org/abs/1810.04576
====================================================
Temporal Cross-Media Retrieval with Soft-Smoothing (David Semedo - 10 October, 2018)
Building on such knowledge, we propose a novel temporal cross-media neural architecture, that departs from standard cross-media methods, by explicitly accounting for the temporal dimension through temporal subspace learning
Link: https://arxiv.org/abs/1810.04547
====================================================
Cutting Throughput on the Edge:App-Aware Placement in Fog Computing (Francescomaria Faticanti - 10 October, 2018)
We consider a scenario where the emerging microservice architecture allows for the design of applications as cascades of coupled microservice modules
Link: https://arxiv.org/abs/1810.04442
====================================================
Global Search with Bernoulli Alternation Kernel for Task-oriented Grasping Informed by Simulation (Rika Antonova - 10 October, 2018)
Our further contribution is a neural network architecture and training pipeline that use experience from grasping objects in simulation to learn grasp stability scores
Link: https://arxiv.org/abs/1810.04438
====================================================
Improving Neural Text Simplification Model with Simplified Corpora (Jipeng Qiang - 10 October, 2018)
Target-side simplified sentences plays an important role in boosting fluency for statistical TS, and we investigate the use of simplified sentences to train, with no changes to the network architecture. We train encoder-decoder model using synthetic sentence pairs and original sentence pairs, which can obtain substantial improvements on the available WikiLarge data and WikiSmall data compared with the state-of-the-art methods.
Link: https://arxiv.org/abs/1810.04428
====================================================
Performance analysis and optimization of the JOREK code for many-core CPUs (T. B. Fehér - 10 October, 2018)
This report investigates the performance of the JOREK code on the Intel Knights Landing and Skylake processor architectures
Link: https://arxiv.org/abs/1810.04413
====================================================
Autonomous Urban Localization and Navigation with Limited Information (Jordan Chipka - 9 October, 2018)
This paper addresses the issue of autonomous driving in an urban environment by investigating algorithms and an architecture to enable fully functional autonomous driving with limited information
Link: https://arxiv.org/abs/1810.04243
====================================================
Using learning to control artificial avatars in human motor coordination tasks (Maria Lombardi - 9 October, 2018)
Then, we embed such synthetic model in a control architecture based on reinforcement learning to synthesize a cyber-agent able to mimic the behaviour of a specific human performing a joint motor task with one or more individuals.
Link: https://arxiv.org/abs/1810.04191
====================================================
Optimized Gated Deep Learning Architectures for Sensor Fusion (Myung Seok Shim - 8 October, 2018)
In this paper, we address several limitations of the baseline negated architecture by proposing two further optimized architectures: a coarser-grained gated architecture employing (feature) group-level fusion weights and a two-stage gated architectures leveraging both the group-level and feature level fusion weights. Using driving mode prediction and human activity recognition datasets, we demonstrate the significant performance improvements brought by the proposed gated architectures and also their robustness in the presence of sensor noise and failures.
Link: https://arxiv.org/abs/1810.04160
====================================================
Conversational Group Detection With Deep Convolutional Networks (Mason Swofford - 7 October, 2018)
Specifically, this novel network builds on the PointNet architecture which allows unordered inputs of variable sizes
Link: https://arxiv.org/abs/1810.04039
====================================================
DeepNIS: Deep Neural Network for Nonlinear Electromagnetic Inverse Scattering (Lianlin Li - 4 October, 2018)
To tackle these difficulties, we, for the first time to our best knowledge, exploit a connection between the deep neural network (DNN) architecture and the iterative method of nonlinear EM inverse scattering
Link: https://arxiv.org/abs/1810.03990
====================================================
Deep Decoder: Concise Image Representations from Untrained Non-convolutional Networks (Reinhard Heckel - 2 October, 2018)
The deep decoder has a simple architecture with no convolutions and fewer weight parameters than the output dimensionality. Further, underparameterization provides a barrier to overfitting, allowing the deep decoder to have state-of-the-art performance for denoising
Link: https://arxiv.org/abs/1810.03982
====================================================
Cognitive Architecture for a Connected World (Shaun C. D&#39;Souza - 26 September, 2018)
The architecture is a super-set to present day infrastructure web computing. The Java virtual machine (JVM) derives from the stack architecture
Link: https://arxiv.org/abs/1810.03955
====================================================
Convolutional Neural Networks In Convolution (Xiaobo Huang - 9 October, 2018)
In contrast, We propose a novel wider Convolutional Neural Networks (CNN) architecture, motivated by the Multi-column Deep Neural Networks and the Network In Network(NIN), aiming for higher accuracy without input data transmutation. In our architecture, namely "CNN In Convolution"(CNNIC), a small CNN, instead of the original generalized liner model(GLM) based filters, is convoluted as kernel on the original image, serving as feature extracting layer of this networks
Link: https://arxiv.org/abs/1810.03946
====================================================
Functionally Modular and Interpretable Temporal Filtering for Robust Segmentation (Jörg Wagner - 9 October, 2018)
Using photorealistic, synthetic video data, we show the ability of the proposed architecture to overcome data-inherent perturbations
Link: https://arxiv.org/abs/1810.03867
====================================================
A software-defined architecture for control of IoT Cyberphysical Systems (Ala&#39; Darabseh - 9 October, 2018)
Based on software-defined principles, we propose a holistic architecture for Cyberphysical Systems (CPS) and Internet of Things (IoT) applications, and highlight the merits pertaining to scalability, flexibility, robustness, interoperability, and cyber security
Link: https://arxiv.org/abs/1810.03822
====================================================
Algorithmic Aspects of Inverse Problems Using Generative Models (Chinmay Hegde - 8 October, 2018)
For certain generative network architectures, we establish a simple non-convex algorithmic approach that (a) theoretically enjoys linear convergence guarantees for certain inverse problems, and (b) empirically improves upon conventional techniques such as back-propagation
Link: https://arxiv.org/abs/1810.03587
====================================================
Meta-Learning: A Survey (Joaquin Vanschoren - 8 October, 2018)
Not only does this dramatically speed up and improve the design of machine learning pipelines or neural architectures, it also allows us to replace hand-engineered algorithms with novel approaches learned in a data-driven way. In this chapter, we provide an overview of the state of the art in this fascinating and continuously evolving field.
Link: https://arxiv.org/abs/1810.03548
====================================================
Security Analysis of Deep Neural Networks Operating in the Presence of Cache Side-Channel Attacks (Sanghyun Hong - 8 October, 2018)
Once the attacker observes function invocations that map directly to architecture attributes of the victim network, the attacker can reconstruct the victim's entire network architecture. Based on the extracted architecture attributes, we also demonstrate that an attacker can build a meta-model that accurately fingerprints the architecture and family of the pre-trained model in a transfer learning setting
Link: https://arxiv.org/abs/1810.03487
====================================================
Handover between Macrocell and Femtocell for UMTS based Networks (Mostafa Zaman Chowdhury - 4 October, 2018)
Some modifications of existing network and protocol architecture for the integration of femtocell networks with the existing UMTS based macrocell networks are essential
Link: https://arxiv.org/abs/1810.03469
====================================================
Probabilistic Argumentation and Information Algebras of Probability Potentials on Families of Compatible Frames (Juerg Kohlas - 8 October, 2018)
As a consequence the well-known local computation architectures of Bayesian networks for inference apply also for the potentials on f.c.f
Link: https://arxiv.org/abs/1810.03423
====================================================
Practical Implementation of Memristor-Based Threshold Logic Gates (Georgios Papandroulidakis - 8 October, 2018)
Current advances in emerging memory technologies enable novel and unconventional computing architectures for high-performance and low-power electronic systems, capable of carrying out massively parallel operations at the edge
Link: https://arxiv.org/abs/1810.03333
====================================================
A Comprehensive Study on Load Balancers for VNF chains Horizontal Scaling (Jiefei Ma - 7 October, 2018)
We conducted extensive simulations to test the feasibility of the architecture and evaluate the performance of our implementation.
Link: https://arxiv.org/abs/1810.03238
====================================================
Rethinking Recurrent Latent Variable Model for Music Composition (Eunjeong Stella Koh - 7 October, 2018)
To generate sequential data, the model uses an encoder-decoder architecture with latent probabilistic connections to capture the hidden structure of music
Link: https://arxiv.org/abs/1810.03226
====================================================
Principled Deep Neural Network Training through Linear Programming (Daniel Bienstock - 7 October, 2018)
In particular, we obtain polynomial time algorithms for training for a given fixed network architecture. Our work applies more broadly to empirical risk minimization problems which allows us to generalize various previous results and obtain new complexity results for previously unstudied architectures in the proper learning setting.
Link: https://arxiv.org/abs/1810.03218
====================================================
Image Completion on CIFAR-10 (Mason Swofford - 7 October, 2018)
This project performed image completion on CIFAR-10, a dataset of 60,000 32x32 RGB images, using three different neural network architectures: fully convolutional networks, convolutional networks with fully connected layers, and encoder-decoder convolutional networks
Link: https://arxiv.org/abs/1810.03213
====================================================
Real-Time Workload Classification during Driving using HyperNetworks (Ruohan Wang - 7 October, 2018)
We also demonstrate the merit of our proposed architecture by showing improved performance over other LSTM-based methods.
Link: https://arxiv.org/abs/1810.03145
====================================================
CSI-Net: Unified Human Body Characterization and Action Recognition (Fei Wang - 6 October, 2018)
To address these issues, we build CSI-Net, a unified Deep Neural Network (DNN), that fully utilizes the strength of deep feature representation and the power of existing DNN architectures for CSI-based human sensing problems
Link: https://arxiv.org/abs/1810.03064
====================================================
Text-based Sentiment Analysis and Music Emotion Recognition (Erion Çano - 6 October, 2018)
This thesis addresses the above problems to provide methodological and practical insights for utilizing neural networks on sentiment analysis of texts and achieving state of the art results. The derived architecture achieves competitive performance on sentiment polarity analysis of movie, business and product reviews.
Link: https://arxiv.org/abs/1810.03031
====================================================
Team NimbRo at MBZIRC 2017: Autonomous Valve Stem Turning using a Wrench (Max Schwarz - 6 October, 2018)
An object detection architecture based on deep neural networks is used to find and select the correct tool from grayscale images
Link: https://arxiv.org/abs/1810.02997
====================================================
Practical Aspects of LTE Network Design and Deployment (Ayman Elnashar - 6 October, 2018)
The end-to-end architecture of the LTE network and different deployment scenarios are presented
Link: https://arxiv.org/abs/1810.02970
====================================================
Co-Stack Residual Affinity Networks with Multi-level Attention Refinement for Matching Text Sequences (Yi Tay - 6 October, 2018)
CSRAN incorporates two novel components to take advantage of the stacked architecture. We conduct extensive experiments on six well-studied text sequence matching datasets, achieving state-of-the-art performance on all.
Link: https://arxiv.org/abs/1810.02938
====================================================
Q-map: a Convolutional Approach for Goal-Oriented Reinforcement Learning (Fabio Pardo - 5 October, 2018)
This architecture is similar to Horde with parameter sharing and allows the agent to discover correlations between visual patterns and navigation
Link: https://arxiv.org/abs/1810.02927
====================================================
Interaction and Autonomy in RoboCup@Home and Building-Wide Intelligence (Justin Hart - 5 October, 2018)
We discuss the contrasts and synergies of these efforts, highlighting how our work to build a RoboCup@Home Domestic Standard Platform League entry led us to identify an integrated software architecture that could support both projects
Link: https://arxiv.org/abs/1810.02919
====================================================
Sifaka: Text Mining Above a Search API (Cameron VandenBerg - 5 October, 2018)
Text mining and analytics software has become popular, but little attention has been paid to the software architectures of such systems
Link: https://arxiv.org/abs/1810.02907
====================================================
Hierarchical Recurrent Filtering for Fully Convolutional DenseNets (Jörg Wagner - 5 October, 2018)
The resulting recurrent architecture temporally filters representations on all abstraction levels in a hierarchical manner, while decoupling temporal dependencies from scene representation
Link: https://arxiv.org/abs/1810.02766
====================================================
ReTiCaM: Real-time Human Performance Capture from Monocular Video (Marc Habermann - 5 October, 2018)
In order to achieve real-time performance of over 25Hz, we design a pipelined parallel architecture using the CPU and two commodity GPUs
Link: https://arxiv.org/abs/1810.02648
====================================================
Game-Based Approach for QoS Provisioning and Interference Management in Heterogeneous Networks (A. S. M. Zadid Shifat - 5 October, 2018)
Future generation wireless networks will use autonomous and distributed architecture for ameliorating the efficacy and flexibility of communication systems
Link: https://arxiv.org/abs/1810.02592
====================================================
Class-Based Interference Management in Wireless Networks (Mohammad Arif Hossain - 5 October, 2018)
The architecture allows improved signal to interference plus noise ratio (SINR) for real-time traffic along with intensification in the bandwidth utilization of the network
Link: https://arxiv.org/abs/1810.02557
====================================================
Radio Resource Management for Dynamic Channel Borrowing Scheme in Wireless Networks (Mohammad Arif Hossain - 5 October, 2018)
In this paper, an efficient dynamic channel borrowing architecture has been proposed that ensures better QoS. If a channel is borrowed from adjacent cells and causing interference, we also propose architecture that diminishes the interference problem
Link: https://arxiv.org/abs/1810.02539
====================================================
An Edge-Computing Based Architecture for Mobile Augmented Reality (Jinke Ren - 4 October, 2018)
Based on the proposed architecture, we further develop an innovated operation mechanism to improve the performance of mobile AR applications. Three key technologies are also discussed to further assist the proposed AR architecture
Link: https://arxiv.org/abs/1810.02509
====================================================
Interference Mitigation Using Dynamic Frequency Re-use for Dense Femtocell Network Architectures (Mostafa Zaman Chowdhury - 4 October, 2018)
For highly dense femtocells, we propose the functionalities of self organizing network (SON) based femtocell network architecture
Link: https://arxiv.org/abs/1810.02486
====================================================
FashionNet: Personalized Outfit Recommendation with Deep Neural Network (Tong He - 4 October, 2018)
We design and compare three alternative architectures for FashionNet
Link: https://arxiv.org/abs/1810.02443
====================================================
Map Memorization and Forgetting in the IARA Autonomous Car (Thomas Teixeira - 4 October, 2018)
The strategy was inspired by an analogy between the memory architecture believed to exist in the human brain and the maps maintained by an autonomous vehicle
Link: https://arxiv.org/abs/1810.02355
====================================================
SNIP: Single-shot Network Pruning based on Connection Sensitivity (Namhoon Lee - 4 October, 2018)
This eliminates the need for both pretraining as well as the complex pruning schedule while making it robust to architecture variations. Our method obtains extremely sparse networks with virtually the same accuracy as the reference network on image classification tasks and is broadly applicable to various architectures including convolutional, residual and recurrent networks
Link: https://arxiv.org/abs/1810.02340
====================================================
A Practical Approach to Sizing Neural Networks (Gerald Friedland - 4 October, 2018)
First, we present four easily applicable rules to analytically determine the capacity of neural network architectures. This allows the comparison of the efficiency of different network architectures independently of a task
Link: https://arxiv.org/abs/1810.02328
====================================================
MyCaffe: A Complete C# Re-Write of Caffe with Reinforcement Learning (David W. Brown - 4 October, 2018)
This article describes the general architecture of MyCaffe including the newly added MyCaffeTrainerRL for Reinforcement Learning
Link: https://arxiv.org/abs/1810.02272
====================================================
A Large-Scale Test Set for the Evaluation of Context-Aware Pronoun Translation in Neural Machine Translation (Mathias Müller - 4 October, 2018)
Our experiments also show the effectiveness of parameter tying for multi-encoder architectures.
Link: https://arxiv.org/abs/1810.02268
====================================================
Italian Event Detection Goes Deep Learning (Tommaso Caselli - 4 October, 2018)
The results also provide further evidence that embeddings have a major impact on the performance of such architectures.
Link: https://arxiv.org/abs/1810.02229
====================================================
Design and Evaluation of A Data Partitioning-Based Intrusion Management Architecture for Database Systems (Muhamad Felemban - 5 October, 2018)
The proposed architecture for PIMS includes novel IB-centric response and recovery mechanisms, which executes compensating transactions
Link: https://arxiv.org/abs/1810.02061
====================================================
Learning Bidirectional LSTM Networks for Synthesizing 3D Mesh Animation Sequences (Yi-Ling Qiao - 3 October, 2018)
Deep architectures that operate directly on mesh sequences remain unexplored, due to the following major barriers: meshes with irregular triangles, sequences containing rich temporal information and flexible deformations
Link: https://arxiv.org/abs/1810.02042
====================================================
Detecting DGA domains with recurrent neural networks and side information (Ryan R. Curtin - 3 October, 2018)
We then describe our new modeling approach, which is a combination of a novel recurrent neural network architecture with domain registration side information. The model's performance compared to the state of the art is best for DGA families that resemble English words
Link: https://arxiv.org/abs/1810.02023
====================================================
LIT: Block-wise Intermediate Representation Training for Model Compression (Animesh Koratana - 1 October, 2018)
We also show that applying LIT to identical student/teacher architectures increases the accuracy of the student model above the teacher model, outperforming the recently-proposed Born Again Networks procedure on ResNet, ResNeXt, and VDCNN
Link: https://arxiv.org/abs/1810.01937
====================================================
Cloud4IoT: a heterogeneous, distributed and autonomic cloud platform for the IoT (Daniele Pizzolli - 3 October, 2018)
Furthermore it shifts traditionally centralized Cloud architectures towards a more distributed and decentralized computation paradigm, as required by IoT technologies, bridging the gap between Cloud Computing and IoT ecosystems. The hierarchical architecture of Cloud4IoThosts a central Cloud platform and multiple remote edge Cloud modules supporting dedicated devices, namely the IoT Gateways, through which new sensor objects are made accessible to the platform
Link: https://arxiv.org/abs/1810.01839
====================================================
Deep Fundamental Matrix Estimation without Correspondences (Omid Poursaeed - 2 October, 2018)
In this paper, we propose novel neural network architectures to estimate fundamental matrices in an end-to-end manner without relying on point correspondences
Link: https://arxiv.org/abs/1810.01575
====================================================
Distributing and Obfuscating Firewalls via Oblivious Bloom Filter Evaluation (Ken Goss - 2 October, 2018)
In this paper, we develop an innovative distributed architecture to obliviously manage and evaluate firewalls to prevent both insider and external attacks oriented to the firewalls
Link: https://arxiv.org/abs/1810.01571
====================================================
FFJORD: Free-form Continuous Dynamics for Scalable Reversible Generative Models (Will Grathwohl - 3 October, 2018)
The result is a continuous-time invertible generative model with unbiased density estimation and one-pass sampling, while allowing unrestricted neural network architectures. We demonstrate our approach on high-dimensional density estimation, image generation, and variational inference, achieving the state-of-the-art among exact likelihood methods with efficient sampling.
Link: https://arxiv.org/abs/1810.01367
====================================================
Learning with Random Learning Rates (Léonard Blier - 3 October, 2018)
Perhaps surprisingly, stochastic gradient descent (SGD) with Alrao performs close to SGD with an optimally tuned learning rate, for various architectures and problems
Link: https://arxiv.org/abs/1810.01322
====================================================
Know What Your Neighbors Do: 3D Semantic Segmentation of Point Clouds (Francis Engelmann - 2 October, 2018)
In this paper, we present a deep learning architecture which addresses the problem of 3D semantic segmentation of unstructured point clouds. We show how to apply these mechanisms to the task of 3D semantic segmentation of point clouds and report state-of-the-art performance on indoor and outdoor datasets.
Link: https://arxiv.org/abs/1810.01151
====================================================
The Dreaming Variational Autoencoder for Reinforcement Learning Environments (Per-Arne Andersen - 2 October, 2018)
This paper presents The Dreaming Variational Autoencoder (DVAE), a neural network based generative modeling architecture for exploration in environments with sparse feedback
Link: https://arxiv.org/abs/1810.01112
====================================================
Future Architecture of the Interplanetary Internet (Ahmad Alhilal - 2 October, 2018)
More specifically, we consider the IPN as the combination of physical infrastructure, network architecture, and technologies to provide communication and navigation services for missions and further applications
Link: https://arxiv.org/abs/1810.01093
====================================================
Improving Sentence Representations with Multi-view Frameworks (Shuai Tang - 2 October, 2018)
Motivated by the asymmetry in the two hemispheres of the human brain as well as the observation that different learning architectures tend to emphasise different aspects of sentence meaning, we present two multi-view frameworks for learning sentence representations in an unsupervised fashion
Link: https://arxiv.org/abs/1810.01064
====================================================
Real-Time Planning with Multi-Fidelity Models for Agile Flights in Unknown Environments (Jesus Tordesillas - 1 October, 2018)
Although the planning methodologies vary from platform to platform, many algorithms adopt a hierarchical planning architecture where a slow, low-fidelity global planner guides a fast, high-fidelity local planner. This novel mapping and planning framework for agile flights is validated in simulation, showing replanning times of 5-40 ms in cluttered environments, a value that is 3-30 times faster than similar state-of-the-art planning algorithms.
Link: https://arxiv.org/abs/1810.01035
====================================================
Bayesian Policy Optimization for Model Uncertainty (Gilwoo Lee - 1 October, 2018)
To address challenges from discretizing the continuous latent parameter space, we propose a policy network architecture that independently encodes the belief distribution from the observable state. Our method significantly outperforms algorithms that address model uncertainty without explicitly reasoning about belief distributions, and is competitive with state-of-the-art Partially Observable Markov Decision Process solvers.
Link: https://arxiv.org/abs/1810.01014
====================================================
Towards a Unified Planner For Socially-Aware Navigation (Santosh Balajee Banisetty - 1 October, 2018)
This paper presents a novel architecture to attain a Unified Socially-Aware Navigation (USAN) and explains its need in Socially Assistive Robotics (SAR) applications. In this work, we propose a novel approach for a unified architecture to SAN that can plan and execute trajectories which are human-friendly for an autonomously sensed interaction context
Link: https://arxiv.org/abs/1810.00966
====================================================
How Powerful are Graph Neural Networks? (Keyulu Xu - 1 October, 2018)
We then develop a simple architecture that is provably the most expressive among the class of GNNs and is as powerful as the Weisfeiler-Lehman graph isomorphism test. We empirically validate our theoretical findings on a number of graph classification benchmarks, and demonstrate that our model achieves state-of-the-art performance.
Link: https://arxiv.org/abs/1810.00826
====================================================
Benchmark Analysis of Representative Deep Neural Network Architectures (Simone Bianco - 1 October, 2018)
To measure the indices we experiment the use of DNNs on two different computer architectures, a workstation equipped with a NVIDIA Titan X Pascal and an embedded system based on a NVIDIA Jetson TX1 board. This study is useful for researchers to have a complete view of what solutions have been explored so far and in which research directions are worth exploring in the future; and for practitioners to select the DNN architecture(s) that better fit the resource constraints of practical deployments and applications
Link: https://arxiv.org/abs/1810.00736
====================================================
Throughput Optimizations for FPGA-based Deep Neural Network Inference (Thorbjörn Posewsky - 28 September, 2018)
In this paper, we propose novel architectures for the inference of previously learned and arbitrary deep neural networks on FPGA-based SoCs that are able to overcome these limitations
Link: https://arxiv.org/abs/1810.00722
====================================================
Perfect Match: A Simple Method for Learning Representations For Counterfactual Inference With Neural Networks (Patrick Schwab - 3 October, 2018)
Here, we present Perfect Match (PM), a method for training neural networks for counterfactual inference that is easy to implement, compatible with any architecture, does not add computational complexity or hyperparameters, and extends to any number of treatments. Our experiments demonstrate that PM outperforms a number of more complex state-of-the-art methods in inferring counterfactual outcomes across several real-world and semi-synthetic datasets.
Link: https://arxiv.org/abs/1810.00656
====================================================
Talaia: a Real time Monitor of Social Media and Digital Press (Iñaki San Vicente - 28 September, 2018)
This paper presents the architecture of the system and describes in detail the different components of the system
Link: https://arxiv.org/abs/1810.00647
====================================================
Learnable Pooling Methods for Video Classification (Sebastian Kmiec - 1 October, 2018)
Rather than using ensembles of existing architectures, we provide an insight on creating new architectures. We obtain testing accuracy similar to the state of the art, while meeting budget constraints, and touch upon strategies to improve the state of the art
Link: https://arxiv.org/abs/1810.00530
====================================================
End-To-End Alzheimer's Disease Diagnosis and Biomarker Identification (Soheil Esmaeilzadeh - 1 October, 2018)
In this paper, we propose a simple 3D Convolutional Neural Networks and exploit its model parameters to tailor the end-to-end architecture for the diagnosis of Alzheimer's disease (AD). Our model can diagnose AD with an accuracy of 94.1\% on the popular ADNI dataset using only MRI data, which outperforms the previous state-of-the-art
Link: https://arxiv.org/abs/1810.00523
====================================================
Benchmarks of ResNet Architecture for Atrial Fibrillation Classification (Roman Khudorozhkov - 30 September, 2018)
In this work we apply variations of ResNet architecture to the task of atrial fibrillation classification
Link: https://arxiv.org/abs/1810.00396
====================================================
Deep, Skinny Neural Networks are not Universal Approximators (Jesse Johnson - 30 September, 2018)
These limitations are typically described in terms of information theoretic bounds, or by comparing the relative complexity needed to approximate example functions between different architectures. In this paper, we examine the topological constraints that the architecture of a neural network imposes on the level sets of all the functions that it is able to approximate
Link: https://arxiv.org/abs/1810.00393
====================================================
Multi-Level Contextual Network for Biomedical Image Segmentation (Amirhossein Dadashzadeh - 30 September, 2018)
We propose a new end-to-end network architecture that effectively integrates local and global contextual patterns of histologic primitives to obtain a more reliable segmentation result
Link: https://arxiv.org/abs/1810.00327
====================================================
Cognitive-LPWAN: Towards Intelligent Wireless Services in Hybrid Low Power Wide Area Networks (Min Chen - 29 September, 2018)
We put forward a cognitive low-power wide-area-network (Cognitive-LPWAN) architecture to safeguard stable and efficient communications in a heterogeneous IoT
Link: https://arxiv.org/abs/1810.00300
====================================================
Modulated Variational auto-Encoders for many-to-many musical timbre transfer (Adrien Bitton - 29 September, 2018)
By further conditioning our system on several different instruments, we can generalize to many-to-many transfer within a single variational architecture able to perform multi-domain transfers. We show that this architecture allows for generative controls in multi-domain transfer, yet remaining light, fast to train and effective on small datasets.
Link: https://arxiv.org/abs/1810.00222
====================================================
Exploiting AWG Free Spectral Range Periodicity in Distributed Multicast Architectures (Kamran Keykhosravi - 29 September, 2018)
Modular optical switch architectures combining wavelength routing based on arrayed waveguide grating (AWG) devices and multicasting based on star couplers hold promise for flexibly addressing the exponentially growing traffic demands in a cost- and power-efficient fashion
Link: https://arxiv.org/abs/1810.00197
====================================================
Knowledge-guided Semantic Computing Network (Guangming Shi - 28 September, 2018)
However, due to the obscured and indescribable black box model of neural networks, it is very difficult to design its architecture, interpret its features and predict its performance
Link: https://arxiv.org/abs/1810.00139
====================================================
Cell Grid Architecture for Maritime Route Prediction on AIS Data Streams (Ciprian Amariei - 28 September, 2018)
Our solution employs a cell grid architecture essentially based on a sequence of hash tables, specifically built for the targeted use case. Moreover, the architecture proposed accommodates also an optionally semi-supervised learning process besides the basic supervised mode.
Link: https://arxiv.org/abs/1810.00090
====================================================
SALSA-TEXT : self attentive latent space based adversarial text generation (Jules Gagnon-Marchand - 8 October, 2018)
In this paper, we take a step to fortify the architectures used in these setups, specifically AAE and ARAE. The experiments demonstrate the proposed (self) attention-based models outperform the state-of-the-art in adversarial code-based text generation.
Link: https://arxiv.org/abs/1809.11155
====================================================
Hierarchical and State-based Architectures for Robot Behavior Planning and Control (Philipp Allgeuer - 28 September, 2018)
In this paper, two behavior control architectures for autonomous agents in the form of cross-platform C++ frameworks are presented, the State Controller Library and the Behavior Control Framework
Link: https://arxiv.org/abs/1809.11067
====================================================
SConE: Siamese Constellation Embedding Descriptor for Image Matching (Tomasz Trzcinski - 28 September, 2018)
To that end, we use a Siamese architecture that learns a low-dimensional feature embedding of keypoint constellation by maximizing the distances between non-corresponding pairs of matched image patches, while minimizing it for correct matches. The 48-dimensional oating point descriptor that we train is built on top of the state-of-the-art FREAK descriptor achieves significant performance improvement over the competitors on a challenging TUM dataset.
Link: https://arxiv.org/abs/1809.11054
====================================================
Real-time Dynamic Object Detection for Autonomous Driving using Prior 3D-Maps (B Ravi Kiran - 28 September, 2018)
We propose a rejection cascade architecture to subtract road regions and other 3D regions separately
Link: https://arxiv.org/abs/1809.11036
====================================================
Domain Generalization with Domain-Specific Aggregation Modules (Antonio D&#39;Innocente - 28 September, 2018)
This paper contributes to this research thread, proposing a deep architecture that maintains separated the information about the available source domains data while at the same time leveraging over generic perceptual information. Experiments on two different benchmark databases show the power of our approach, reaching the new state of the art in domain generalization.
Link: https://arxiv.org/abs/1809.10966
====================================================
Blockchain and Smart-contracts Modeled in a SwarmESB Ecosystem (Ioana Stanescu - 28 September, 2018)
Since the appearance of Bitcoin, Blockchain has known many implementations based on P2P architectures
Link: https://arxiv.org/abs/1809.10915
====================================================
Comparison of Self-Aware and Organic Computing Systems (Najma Gill - 28 August, 2018)
This paper is written to compare self-awareness and organic computing by discussing their definitions, properties and architecture.
Link: https://arxiv.org/abs/1809.10846
====================================================
Inverse Transport Networks (Chengqian Che - 27 September, 2018)
We introduce inverse transport networks as a learning architecture for inverse rendering problems where, given input image measurements, we seek to infer physical scene parameters such as shape, material, and illumination
Link: https://arxiv.org/abs/1809.10820
====================================================
Visual Analytics for Automated Model Discovery (Dylan Cashman - 2 October, 2018)
First, we propose an architecture based on an extension of existing visual analytics frameworks. Then we describe a prototype system Snowcat, developed according to the presented framework and architecture, that aids users in generating models for a diverse set of data and modeling tasks.
Link: https://arxiv.org/abs/1809.10782
====================================================
Learning to Train a Binary Neural Network (Joseph Bethge - 27 September, 2018)
Within this framework, we systematically evaluated different network architectures and hyperparameters to provide useful insights on how to train a binary neural network
Link: https://arxiv.org/abs/1809.10463
====================================================
Sample Efficient Adaptive Text-to-Speech (Yutian Chen - 27 September, 2018)
We introduce and benchmark three strategies: (i) learning the speaker embedding while keeping the WaveNet core fixed, (ii) fine-tuning the entire architecture with stochastic gradient descent, and (iii) predicting the speaker embedding with a trained neural network encoder. The experiments show that these approaches are successful at adapting the multi-speaker neural network to new speakers, obtaining state-of-the-art results in both sample naturalness and voice similarity with merely a few minutes of audio data from new speakers.
Link: https://arxiv.org/abs/1809.10460
====================================================
Wafer Quality Inspection using Memristive LSTM, ANN, DNN and HTM (Kazybek Adam - 27 September, 2018)
The application of non-volatile memristive devices in these architectures ensures low power consumption, small on-chip area scalability. We demonstrate that Long-Short Term Memory (LSTM) outperforms other architectures for the same number of training iterations, and has relatively low on-chip area and power consumption.
Link: https://arxiv.org/abs/1809.10438
====================================================
Making View Update Strategies Programmable - Toward Controlling and Sharing Distributed Data - (Yasuhito Asano - 27 September, 2018)
We show that Datalog can be used for specifying view update strategies whose unique view definition can be automatically derived, present a novel P2P-based programmable architecture for distributed data management where updatable views are fully utilized for controlling and sharing distributed data, and demonstrate its usefulness through the development of a privacy-preserving ride-sharing alliance system.
Link: https://arxiv.org/abs/1809.10357
====================================================
Geometry-Aware Network for Non-Rigid Shape Prediction from a Single View (Albert Pumarola - 26 September, 2018)
We train this architecture in an end-to-end manner using a large dataset of synthetic renderings of shapes under different levels of deformation, material properties, textures and lighting conditions. We evaluate our approach on a test split of this dataset and available real benchmarks, consistently improving state-of-the-art solutions with a significantly lower computational time.
Link: https://arxiv.org/abs/1809.10305
====================================================
Autonomously and Simultaneously Refining Deep Neural Network Parameters by a Bi-Generative Adversarial Network Aided Genetic Algorithm (Yantao Lu - 24 September, 2018)
The choice of parameters, and the design of the network architecture are important factors affecting the performance of deep neural networks. Our proposed approach can be used to autonomously refine the number of convolutional layers and dense layers, number and size of kernels, and the number of neurons for the dense layers; choose the type of the activation function; and decide whether to use dropout and batch normalization or not, to improve the accuracy of different deep neural network architectures
Link: https://arxiv.org/abs/1809.10244
====================================================
Leveraging Transfer Learning for Segmenting Lesions and their Attributes in Dermoscopy Images (Navid Alemi Koohbanani - 23 September, 2018)
The proposed framework is inspired by the well-known UNet architecture
Link: https://arxiv.org/abs/1809.10243
====================================================
Generic Vehicle Tracking Framework Capable of Handling Occlusions Based on Modified Mixture Particle Filter (Jiachen Li - 20 September, 2018)
To overcome this challenge, we propose a generic vehicle tracking framework based on modified mixture particle filter, which can make the number of tracking targets adaptive to real-time observations and track all the vehicles within sensor range simultaneously in a uniform architecture without explicit data association
Link: https://arxiv.org/abs/1809.10237
====================================================
Brokering Policies and Execution Monitors for IoT Middleware (Juan Carlos Fuentes Carranza - 27 September, 2018)
In this work, we propose to extend the modular event-based system architecture of Fiege et al., to incorporate brokering policies and execution monitors, in order to address the three protection challenges mentioned above
Link: https://arxiv.org/abs/1809.10134
====================================================
Language Modeling Teaches You More Syntax than Translation Does: Lessons Learned Through Auxiliary Task Analysis (Kelly W. Zhang - 26 September, 2018)
We make a fair comparison between the tasks by holding constant the quantity and genre of the training data, as well as the LSTM architecture
Link: https://arxiv.org/abs/1809.10040
====================================================
Learning through Probing: a decentralized reinforcement learning architecture for social dilemmas (Nicolas Anastassacos - 26 September, 2018)
Multi-agent reinforcement learning has received significant interest in recent years notably due to the advancements made in deep reinforcement learning which have allowed for the developments of new architectures and learning algorithms. Using social dilemmas as the training ground, we present a novel learning architecture, Learning through Probing (LTP), where agents utilize a probing mechanism to incorporate how their opponent's behavior changes when an agent takes an action
Link: https://arxiv.org/abs/1809.10007
====================================================
Knowledge extraction, modeling and formalization: EEG case study (Dmitry Morozov - 11 September, 2018)
The contribution of this paper lies in the specification of desritizition procedure and the architecture of FCA experiment
Link: https://arxiv.org/abs/1809.09955
====================================================
GPU Accelerated Similarity Self-Join for Multi-Dimensional Data (Michael Gowanlock - 26 September, 2018)
Across most scenarios on real-world and synthetic datasets, our algorithm outperforms the parallel state-of-the-art approach. Exascale systems are converging on heterogeneous distributed-memory architectures
Link: https://arxiv.org/abs/1809.09930
====================================================
Coordinating and Integrating Faceted Classification with Rich Semantic Modeling (Robert B. Allen - 25 September, 2018)
We compare a faceted thesaurus, the Art and Architecture Thesaurus (AAT), with ontologies derived from the Basic Formal Ontology (BFO2), which is an upper (or formal) ontology widely used to describe entities in biomedicine
Link: https://arxiv.org/abs/1809.09548
====================================================
Network Coding for Critical Infrastructure Networks (Rakesh Kumar - 25 September, 2018)
To that end, we propose an architecture that realizes linear NC in programmable networks by decomposing the linear NC functions into the atomic coding primitives
Link: https://arxiv.org/abs/1809.09487
====================================================
Defining the Collective Intelligence Supply Chain (Iain Barclay - 25 September, 2018)
A proposal is made for the development of a supply chain model for tracking the creation and use of crowdsourced collective intelligence assets, with a blockchain based decentralised architecture identified as an appropriate means of providing validation, accountability and fairness.
Link: https://arxiv.org/abs/1809.09444
====================================================
Memory Management in Successive-Cancellation based Decoders for Multi-Kernel Polar Codes (Valerio Bioglio - 25 September, 2018)
It can be used on many types of hardware architectures and different flavors of SC decoding algorithms
Link: https://arxiv.org/abs/1809.09436
====================================================
A Framework for Data-Driven Physical Security and Insider Threat Detection (Vasileios Mavroeidis - 25 September, 2018)
Moreover, the information can be used to enrich the available intelligence (about intrusion attempts) that can form use cases to detect and remediate limitations in the system, such as loosely-coupled provenance graphs that in many cases indicate weaknesses in the physical security architecture
Link: https://arxiv.org/abs/1809.09434
====================================================
Supervised Neural Models Revitalize the Open Relation Extraction (Shengbin Jia - 25 September, 2018)
Although a few people try to use neural network based models to improve the ORE task performance recently, it is always intractable for ORE to produce supervised systems based on various neural architectures
Link: https://arxiv.org/abs/1809.09408
====================================================
A 5G Architecture for The Factory of the Future (Stephan Ludwig - 25 September, 2018)
Based on these use cases and the ones by 3GPP, a 5G system architecture for the factory of the future is proposed
Link: https://arxiv.org/abs/1809.09396
====================================================
HSTREAM: A directive-based language extension for heterogeneous stream computing (Suejb Memeti - 25 September, 2018)
Programming such systems require advanced knowledge of several hardware architectures and device-specific programming models, including OpenMP and CUDA
Link: https://arxiv.org/abs/1809.09387
====================================================
Hierarchical Deep Multiagent Reinforcement Learning (Hongyao Tang - 25 September, 2018)
Several hierarchical deep MARL architectures, i.e., Ind-hDQN, hCom and hQmix, are introduced for different learning paradigms
Link: https://arxiv.org/abs/1809.09332
====================================================
No Multiplication? No Floating Point? No Problem! Training Networks for Efficient Inference (Shumeet Baluja - 28 September, 2018)
The memory needed to deploy our discretized networks is less than one third of the equivalent architecture that does use floating-point operations.
Link: https://arxiv.org/abs/1809.09244
====================================================
SDN-Based Resource Management for Autonomous Vehicular Networks: A Multi-Access Edge Computing Approach (Haixia Peng - 24 September, 2018)
In this article, a new architecture that combines multi-access edge computing (MEC) and software-defined networking (SDN) is proposed for flexible resource management and enhanced resource utilization
Link: https://arxiv.org/abs/1809.08966
====================================================
Representing Sets as Summed Semantic Vectors (Douglas Summers-Stay - 24 September, 2018)
Representing meaning in the form of high dimensional vectors is a common and powerful tool in biologically inspired architectures
Link: https://arxiv.org/abs/1809.08823
====================================================
Data and Spectrum Trading Policies in a Trusted Cognitive Dynamic Network (B. Lorenzo - 24 September, 2018)
In this paper, we consider a cognitive dynamic network architecture (CDNA) where primary users (PUs) are rewarded for sharing their connectivities and acting as access points for secondary users (SUs)
Link: https://arxiv.org/abs/1809.08805
====================================================
An Architecture for Person-Following using Active Target Search (Minkyu Kim - 24 September, 2018)
This paper addresses a novel architecture for person-following robots using active search. Using Toyota Human Support Robot (HSR), real-time experiments demonstrate that the proposed architecture can generate fast, efficient person-following behaviors.
Link: https://arxiv.org/abs/1809.08793
====================================================
Deformable Stacked Structure for Named Entity Recognition (Shuyang Cao - 28 September, 2018)
Currently, the dominating architecture consists of a bi-directional recurrent neural network (RNN) as the encoder and a conditional random field (CRF) as the decoder. Our model achieves the state-of-the-art performances on the OntoNotes dataset.
Link: https://arxiv.org/abs/1809.08730
====================================================
Deep Knowledge Tracing and Dynamic Student Classification for Knowledge Tracing (Sein Minn - 23 September, 2018)
In this paper, we propose a novel model for knowledge tracing that i) captures students' learning ability and dynamically assigns students into distinct groups with similar ability at regular time intervals, and ii) combines this information with a Recurrent Neural Network architecture known as Deep Knowledge Tracing. Experimental results confirm that the proposed model is significantly better at predicting student performance than well known state-of-the-art techniques for student modelling.
Link: https://arxiv.org/abs/1809.08713
====================================================
Unsupervised Learning of Dense Optical Flow and Depth from Sparse Event Data (Chengxi Ye - 23 September, 2018)
To tackle this low level vision task, we use a novel encoder-decoder neural network architecture that aggregates multi-level features and addresses the problem at multiple resolutions
Link: https://arxiv.org/abs/1809.08625
====================================================
Shift-based Primitives for Efficient Convolutional Neural Networks (Huasong Zhong - 24 September, 2018)
We blend these shift-based primitives with point-wise group convolution and built two inference-efficient CNN architectures named AddressNet and Enhanced AddressNet
Link: https://arxiv.org/abs/1809.08458
====================================================
Focus On What's Important: Self-Attention Model for Human Pose Estimation (Guanxiong Sun - 21 September, 2018)
In this work, we proposed a convolutional network architecture combined with the novel attention model. It's a multi-stage architecture
Link: https://arxiv.org/abs/1809.08371
====================================================
Temporal Interpolation as an Unsupervised Pretraining Task for Optical Flow Estimation (Jonas Wulff - 21 September, 2018)
Using this unsupervised pre-training, our network outperforms similar architectures that were trained supervised using synthetic optical flow.
Link: https://arxiv.org/abs/1809.08317
====================================================
SCOPE: C3SR Systems Characterization and Benchmarking Framework (Carl Pearson - 18 September, 2018)
The Scope project aims to lower the barrier to entry for developing performance benchmarks by providing a software architecture that allows benchmarks to be developed independently, by providing useful C/C++ abstractions and utilities, and by providing a Python package for generating publication-quality plots of resulting measurements.
Link: https://arxiv.org/abs/1809.08311
====================================================
Exclusive Independent Probability Estimation using Deep 3D Fully Convolutional DenseNets for IsoIntense Infant Brain MRI Segmentation (Seyed Raein Hashemi - 27 September, 2018)
We present a 3D FC-DenseNet architecture, an exclusive multilabel patchwise training technique with balanced similarity loss functions and a patch prediction fusion strategy that can be used on new classification and segmentation applications with two or more very similar classes. This strategy improves the training process by reducing its complexity while providing a trained model that works for any size input and is fast and more accurate than many state-of-the-art methods.
Link: https://arxiv.org/abs/1809.08168
====================================================
Towards the Development of a Rule-based Drought Early Warning Expert Systems using Indigenous Knowledge (A. K. Akanbi - 19 September, 2018)
The architecture and components such as knowledge base, JESS inference engine and model base of the system and their functions are presented.
Link: https://arxiv.org/abs/1809.08101
====================================================
Multimodal Dual Attention Memory for Video Story Question Answering (Kyung-Min Kim - 21 September, 2018)
We propose a video story question-answering (QA) architecture, Multimodal Dual Attention Memory (MDAM). For both datasets, MDAM achieves new state-of-the-art results with significant margins compared to the runner-up models
Link: https://arxiv.org/abs/1809.07999
====================================================
Biological plausibility and stochasticity in scalable VO2 active memristor neurons (Wei Yi - 20 September, 2018)
Neuromorphic networks of artificial neurons and synapses can solve computational hard problems with energy efficiencies unattainable for von Neumann architectures
Link: https://arxiv.org/abs/1809.07867
====================================================
Enabling Ultra-Low Delay Teleorchestras using Software Defined Networking (Emmanouil Lakiotakis - 29 August, 2018)
The present work proposes a novel architecture for NMP systems, where the key-innovation is the close collaboration between the network and the application
Link: https://arxiv.org/abs/1809.07864
====================================================
Recurrent Neural Networks based Obesity Status Prediction Using Activity Data (Qinghan Xue - 20 September, 2018)
Meanwhile, the lack of understanding of why those learning models are effective also limits further improvements on their architectures. Thus, in this work, we develop a RNN based time-aware architecture to tackle the challenging problem of handling irregular observation times and relevant feature extractions from longitudinal patient records for obesity status improvement prediction
Link: https://arxiv.org/abs/1809.07828
====================================================
PriPeARL: A Framework for Privacy-Preserving Analytics and Reporting at LinkedIn (Krishnaram Kenthapadi - 20 September, 2018)
We describe the overall design and architecture, and the key modeling components, focusing on the unique challenges associated with privacy, coverage, utility, and consistency
Link: https://arxiv.org/abs/1809.07754
====================================================
Chorus: Differential Privacy via Query Rewriting (Noah Johnson - 23 September, 2018)
We present Chorus, a system with a novel architecture for providing differential privacy for statistical SQL queries
Link: https://arxiv.org/abs/1809.07750
====================================================
On the Fly Orchestration of Unikernels: Tuning and Performance Evaluation of Virtual Infrastructure Managers (Pier Luigi Ventre - 17 September, 2018)
Network Functions Virtualization (NFV) and Cloud Computing are emerging as key trends of 5G network architectures, providing flexibility, fast instantiation times, support of Commercial Off The Shelf hardware and significant cost savings. This functionality is performed by the Virtual Infrastructure Manager (VIM) in the NFV MANagement and Orchestration (MANO) reference architecture
Link: https://arxiv.org/abs/1809.07701
====================================================
Response Time Bounds for Typed DAG Parallel Tasks on Heterogeneous Multi-cores (Meiling Han - 7 August, 2018)
Heterogeneous multi-cores utilize the strength of different architectures for executing particular types of workload, and usually offer higher performance and energy efficiency
Link: https://arxiv.org/abs/1809.07689
====================================================
Next Stop "NoOps": Enabling Cross-System Diagnostics Through Graph-based Composition of Logs and Metrics (Michał Zasadziński - 1 August, 2018)
Systems and their architecture change very rapidly in response to business and user demand
Link: https://arxiv.org/abs/1809.07687
====================================================
AutoAccel: Automated Accelerator Generation and Optimization with Composable, Parallel and Pipeline Architecture (Jason Cong - 30 July, 2018)
CPU-FPGA heterogeneous architectures are attracting ever-increasing attention in an attempt to advance computational capabilities and energy efficiency in today's datacenters. These architectures provide programmers with the ability to reprogram the FPGAs for flexible acceleration of many workloads
Link: https://arxiv.org/abs/1809.07683
====================================================
Investigating Linguistic Pattern Ordering in Hierarchical Natural Language Generation (Shang-Yu Su - 19 September, 2018)
However, such simple encoder-decoder architecture usually fail to generate complex and long sentences, because the decoder has difficulty learning all grammar and diction knowledge well
Link: https://arxiv.org/abs/1809.07629
====================================================
Machine Learning for semi linear PDEs (Quentin Chan-Wai-Nam - 20 September, 2018)
Recent machine learning algorithms dedicated to solving semi-linear PDEs are improved by using different neural network architectures and different parameterizations
Link: https://arxiv.org/abs/1809.07609
====================================================
Learning a Local Feature Descriptor for 3D LiDAR Scans (Ayush Dewan - 20 September, 2018)
Our proposed architecture consists of a Siamese network for learning a feature descriptor and a metric learning network for matching the descriptors
Link: https://arxiv.org/abs/1809.07494
====================================================
Deep Generative Classifiers for Thoracic Disease Diagnosis with Chest X-ray Images (Chengsheng Mao - 19 September, 2018)
In this paper, to make a deep architecture more robust to noise and to reduce overfitting, we propose using deep generative classifiers to automatically diagnose thorax diseases from the chest X-ray images. We implemented our deep generative classifiers based on a number of well-known deterministic neural network architectures, and tested our models on the chest X-ray14 dataset
Link: https://arxiv.org/abs/1809.07436
====================================================
Towards Accountable AI: Hybrid Human-Machine Analyses for Characterizing System Failure (Besmira Nushi - 19 September, 2018)
Pandora leverages both human and system-generated observations to summarize conditions of system malfunction with respect to the input content and system architecture
Link: https://arxiv.org/abs/1809.07424
====================================================
Deep Part Induction from Articulated Object Pairs (Li Yi - 19 September, 2018)
To achieve optimal performance, our architecture alternates between correspondence, deformation flow, and segmentation prediction iteratively in an ICP-like fashion. Our results demonstrate that our method significantly outperforms state-of-the-art techniques in the task of discovering articulated parts of objects
Link: https://arxiv.org/abs/1809.07417
====================================================
A Dataset for Document Grounded Conversations (Kangyan Zhou - 19 September, 2018)
We describe two neural architectures that provide benchmark performance on the task of generating the next response
Link: https://arxiv.org/abs/1809.07358
====================================================
Evolving SDN for Low-Power IoT Networks (Michael Baddeley - 19 September, 2018)
This paper addresses the challenge of bringing high-overhead SDN architecture to IEEE 802.15.4 networks. We consequently introduce μSDN, a lightweight SDN framework for Contiki, with both IPv6 and underlying routing protocol interoperability, as well as optimizing a number of elements within the SDN architecture to reduce control overhead to practical levels
Link: https://arxiv.org/abs/1809.07296
====================================================
Interpretable Textual Neuron Representations for NLP (Nina Poerner - 19 September, 2018)
The representations highlight differences in syntax awareness between the language and visual models of the Imaginet architecture.
Link: https://arxiv.org/abs/1809.07291
====================================================
Modeling Online Discourse with Coupled Distributed Topics (Akshay Srivatsan - 1 October, 2018)
Our model (1) captures discursive interactions along observed reply links in addition to traditional topic information, and (2) incorporates latent distributed representations arranged in a deep architecture, which enables a GPU-based mean-field inference procedure that scales efficiently to large data
Link: https://arxiv.org/abs/1809.07282
====================================================
LFRic: Meeting the challenges of scalability and performance portability in Weather and Climate models (S. V. Adams - 19 September, 2018)
An application called PSyclone, developed at the STFC Hartree centre, can generate the parallel code enabling deployment of a single source science code onto different machine architectures
Link: https://arxiv.org/abs/1809.07267
====================================================
Pose Estimation for Non-Cooperative Spacecraft Rendezvous Using Convolutional Neural Networks (Sumant Sharma - 19 September, 2018)
The primary contribution to the state-of-the-art of this work is the introduction of a novel pose determination method based on Convolutional Neural Networks (CNN) to provide an initial guess of the pose in real-time on-board. Through metrics measuring classification and pose accuracy, it is shown that the presented architecture has desirable robustness and scalable properties.
Link: https://arxiv.org/abs/1809.07238
====================================================
3D Human Pose Estimation with Siamese Equivariant Embedding (Márton Véges - 19 September, 2018)
We propose a siamese architecture that learns a rotation equivariant hidden representation to reduce the need for data augmentation. It achieves state-of-the-art cross-camera error rate among algorithms that use estimated 2D joint coordinates only.
Link: https://arxiv.org/abs/1809.07217
====================================================
Dual Reconstruction Nets for Image Super-Resolution with Gradient Sensitive Loss (Yong Guo - 19 September, 2018)
However, most deep learning methods employ feed-forward architectures, and thus the dependencies between LR and HR images are not fully exploited, leading to limited learning performance
Link: https://arxiv.org/abs/1809.07099
====================================================
Counting the uncountable: deep semantic density estimation from Space (Andres C. Rodriguez - 20 September, 2018)
Experiments also suggest that we need specific CNN architectures in remote sensing instead of blindly applying existing ones from computer vision.
Link: https://arxiv.org/abs/1809.07091
====================================================
NICT's Neural and Statistical Machine Translation Systems for the WMT18 News Translation Task (Benjamin Marie - 19 September, 2018)
Our NMT systems were trained with the transformer architecture using the provided parallel data enlarged with a large quantity of back-translated monolingual data that we generated with a new incremental training framework
Link: https://arxiv.org/abs/1809.07037
====================================================
Removing the Feature Correlation Effect of Multiplicative Noise (Zijun Zhang - 19 September, 2018)
Multiplicative noise, including dropout, is widely used to regularize deep neural networks (DNNs), and is shown to be effective in a wide range of architectures and tasks
Link: https://arxiv.org/abs/1809.07023
====================================================
Decentralized P2P Energy Trading under Network Constraints in a Low-Voltage Network (Jaysson Guerrero - 18 September, 2018)
The results show that our method ensures that energy is exchanged between users under the P2P scheme without violating the network constraints, and that users can still capture the economic benefits of the P2P architecture.
Link: https://arxiv.org/abs/1809.06976
====================================================
On the Learning Dynamics of Deep Neural Networks (Remi Tachet des Combes - 18 September, 2018)
Extending existing results from the linear case, we confirm empirical observations by proving that the classification error also follows a sigmoidal shape in nonlinear architectures
Link: https://arxiv.org/abs/1809.06848
====================================================
Bidirectional Attentional Encoder-Decoder Model and Bidirectional Beam Search for Abstractive Summarization (Kamal Al-Sabahi - 18 September, 2018)
The model has a bidirectional encoder-decoder architecture; in which the encoder and the decoder are bidirectional LSTMs. Experimental results on CNN / Daily Mail dataset show that the proposed model outperforms the current abstractive state-of-the-art models by a considerable mar-gin.
Link: https://arxiv.org/abs/1809.06662
====================================================
Lung Cancer Concept Annotation from Spanish Clinical Narratives (Marjan Najafabadipour - 18 September, 2018)
Therefore, in this paper, we analyze the design of annotators for specific lung cancer concepts that can be integrated over Apache Unstructured Information Management Architecture (UIMA) framework
Link: https://arxiv.org/abs/1809.06639
====================================================
Isolating SDN Control Traffic with Layer-2 Slicing in 6TiSCH Industrial IoT Networks (Michael Baddeley - 18 September, 2018)
However, implementing a centralized SDN architecture in a Low Power and Lossy Network (LLN) faces considerable challenges: not only is controller traffic subject to jitter due to unreliable links and network contention, but the overhead generated by SDN can severely affect the performance of other traffic
Link: https://arxiv.org/abs/1809.06624
====================================================
Towards Deep and Representation Learning for Talent Search at LinkedIn (Rohan Ramanath - 17 September, 2018)
Finally, we present offline and online evaluation results for LinkedIn talent search and recommendation systems, and discuss potential challenges along the path to a fully deep model architecture
Link: https://arxiv.org/abs/1809.06473
====================================================
Segmenting root systems in X-ray computed tomography images using level sets (Amy Tabb - 17 September, 2018)
The segmentation of plant roots from soil and other growing media in X-ray computed tomography images is needed to effectively study the root system architecture without excavation. Our method compares favorably to a state-of-the-art method for root segmentation in X-ray CT image volumes.
Link: https://arxiv.org/abs/1809.06398
====================================================
Déjà Vu: an empirical evaluation of the memorization properties of ConvNets (Alexandre Sablayrolles - 17 September, 2018)
To support our findings, we conduct large-scale experiments on Imagenet and subsets of YFCC-100M with modern architectures such as VGG and Resnet.
Link: https://arxiv.org/abs/1809.06396
====================================================
AUEB at BioASQ 6: Document and Snippet Retrieval (Georgios-Ioannis Brokos - 15 September, 2018)
Our models use novel extensions to deep learning architectures that operate solely over the text of the query and candidate document/snippets
Link: https://arxiv.org/abs/1809.06366
====================================================
Retrospective correction of Rigid and Non-Rigid MR motion artifacts using GANs (Karim Armanious - 6 October, 2018)
MedGAN utilizes a unique combination of non-adversarial losses and a new generator architecture to capture the textures and fine-detailed structures of the desired artifact-free MR images
Link: https://arxiv.org/abs/1809.06276
====================================================
Learning Effective RGB-D Representations for Scene Recognition (Xinhang Song - 17 September, 2018)
Our video recognition architecture combines convolutional and recurrent neural networks (RNNs) that are trained in three steps with increasingly complex data to learn effective features (i.e. Our approach obtains state-of-the-art performances on RGB-D image (NYUD2 and SUN RGB-D) and video (ISIA RGB-D) scene recognition.
Link: https://arxiv.org/abs/1809.06269
====================================================
Vision-based Teleoperation of Shadow Dexterous Hand using End-to-End Deep Neural Network (Shuang Li - 17 September, 2018)
In this paper, we present TeachNet, a novel neural network architecture for intuitive and markerless vision-based teleoperation of dexterous robotic hands. Imitation experiments and grasp tasks teleoperated by novice users demonstrate that TeachNet is more reliable and faster than the state-of-the-art vision-based teleoperation method.
Link: https://arxiv.org/abs/1809.06268
====================================================
Linear and Deformable Image Registration with 3D Convolutional Neural Networks (Stergios Christodoulidis - 13 September, 2018)
Inspired by the recent advances in deep learning, we propose in this paper, a novel convolutional neural network architecture that couples linear and deformable registration within a unified architecture endowed with near real-time performance. We evaluate the performance of our network on the challenging problem of MRI lung registration, and demonstrate superior performance with respect to state of the art elastic registration methods
Link: https://arxiv.org/abs/1809.06226
====================================================
Learning of Multi-Context Models for Autonomous Underwater Vehicles (Bilal Wehbe - 17 September, 2018)
We implement an architecture based on long-short-term-memory (LSTM) networks to learn the different contexts directly from the data
Link: https://arxiv.org/abs/1809.06179
====================================================
Graph Neural Networks for IceCube Signal Classification (Nicholas Choma - 17 September, 2018)
We demonstrate the effectiveness of our GNN architecture on a task classifying IceCube events, where it outperforms both a traditional physics-based method as well as classical 3D convolution neural networks.
Link: https://arxiv.org/abs/1809.06166
====================================================
Periocular Recognition Using CNN Features Off-the-Shelf (Kevin Hernandez-Diaz - 17 September, 2018)
Here, we apply existing pre-trained architectures, proposed in the context of the ImageNet Large Scale Visual Recognition Challenge, to the task of periocular recognition
Link: https://arxiv.org/abs/1809.06157
====================================================
A Deep Learning Framework for Unsupervised Affine and Deformable Image Registration (Bob D. de Vos - 17 September, 2018)
By stacking multiple of these ConvNets into a larger architecture, we are able to perform coarse-to-fine image registration
Link: https://arxiv.org/abs/1809.06130
====================================================
DeepDrum: An Adaptive Conditional Neural Network (Dimos Makris - 17 September, 2018)
Considering music as a sequence of events with multiple complex dependencies, the Long Short-Term Memory (LSTM) architecture has proven very efficient in learning and reproducing musical styles
Link: https://arxiv.org/abs/1809.06127
====================================================
The Impact of On-chip Communication on Memory Technologies for Neuromorphic Systems (Saber Moradi - 1 October, 2018)
Emergent nanoscale non-volatile memory technologies with high integration density offer a promising solution to overcome the scalability limitations of CMOS-based neural networks architectures, by efficiently exhibiting the key principle of neural computation. In this paper, we elaborate on the communication requirements of large-scale neuromorphic designs, and point out the differences with the conventional network-on-chip architectures
Link: https://arxiv.org/abs/1809.06016
====================================================
Systems of bounded rational agents with information-theoretic constraints (Sebastian Gottwald - 16 September, 2018)
Our results suggest that hierarchical architectures of specialized units at lower levels that are coordinated by units at higher levels are optimal, given that each unit's information-processing capability is limited and conforms to constraints on complexity costs.
Link: https://arxiv.org/abs/1809.05897
====================================================
Semantic Interoperability Middleware Architecture for Heterogeneous Environmental Data Sources (A. K. Akanbi - 16 September, 2018)
In this paper, we describe and evaluate a cross-domain middleware architecture that semantically integrates and generate inference from heterogeneous data sources
Link: https://arxiv.org/abs/1809.05890
====================================================
Inspiration Learning through Preferences (Nir Baram - 16 September, 2018)
Unlike classic actor-critic architectures, the critic we use consists of two parts: a) a state-value estimation as in common actor-critic algorithms and b) a single step reward function derived from an expert/agent classifier
Link: https://arxiv.org/abs/1809.05872
====================================================
In Defense of the Classification Loss for Person Re-Identification (Yao Zhai - 16 September, 2018)
We argue that learning global features with classification loss could achieve the same goal, even with a simple and cost-effective architecture design. In extensive experiments, our network outperforms state-of-the-art person re-id frameworks in terms of both accuracy and inference cost.
Link: https://arxiv.org/abs/1809.05864
====================================================
f-VAEs: Improve VAEs with Conditional Flows (Jianlin Su - 16 September, 2018)
Compared with flow-based models such as Glow, f-VAE is more lightweight and converges faster, achieving the same performance under smaller-size architecture.
Link: https://arxiv.org/abs/1809.05861
====================================================
Exploiting Errors for Efficiency: A Survey from Circuits to Algorithms (Phillip Stanley-Marbell - 16 September, 2018)
We present, for the first time, a synthesis of research results on computing systems that only make as many errors as their users can tolerate, from across the disciplines of computer aided design of circuits, digital system design, computer architecture, programming languages, operating systems, and information theory.
Link: https://arxiv.org/abs/1809.05859
====================================================
Low synchronization GMRES algorithms (Kasia Swirydowicz - 15 September, 2018)
Communication-avoiding and pipelined variants of Krylov solvers are critical for the scalability of linear system solvers on future exascale architectures
Link: https://arxiv.org/abs/1809.05805
====================================================
Finding the way from ä to a: Sub-character morphological inflection for the SIGMORPHON 2018 Shared Task (Fynn Schröder - 15 September, 2018)
We propose a neural architecture based on the concepts of UZH (Makarov et al., 2017), adding new ideas and techniques to their key concept and evaluating different combinations of parameters
Link: https://arxiv.org/abs/1809.05742
====================================================
Distinguished Capabilities of Artificial Intelligence Wireless Communication Systems (Xiaohu Ge - 15 September, 2018)
Considering the revolutionary impact of AI technologies on the data, transmission and protocol architecture of wireless communication systems, the future challenges of AI wireless communication systems are analyzed
Link: https://arxiv.org/abs/1809.05673
====================================================
Towards Better Interpretability in Deep Q-Networks (Raghuram Mandyam Annasamy - 14 September, 2018)
In this paper we propose an interpretable neural network architecture for Q-learning which provides a global explanation of the model's behavior using key-value memories, attention and reconstructible embeddings. With a directed exploration strategy, our model can reach training rewards comparable to the state-of-the-art deep Q-learning models
Link: https://arxiv.org/abs/1809.05630
====================================================
Deep Reinforcement Learning Based Mode Selection and Resource Management for Green Fog Radio Access Networks (Yaohua Sun - 14 September, 2018)
Fog radio access networks (F-RANs) are seen as potential architectures to support services of internet of things by leveraging edge caching and edge computing
Link: https://arxiv.org/abs/1809.05629
====================================================
A New Secure Network Architecture to Increase Security Among Virtual Machines in Cloud Computing (Zakaria El Mrabet - 13 September, 2018)
The main purpose of this paper is to present a new secure architecture of Virtual Network machines in order to increase security among virtual machines in a virtualized environment (Xen as a case study). Finally, we present in details new secure architecture and demonstrate how it can face the main security network attacks.
Link: https://arxiv.org/abs/1809.05528
====================================================
Extending Neural Generative Conversational Model using External Knowledge Sources (Prasanna Parthasarathi - 14 September, 2018)
This work proposes an architecture to incorporate unstructured knowledge sources to enhance the next utterance prediction in chit-chat type of generative dialogue models
Link: https://arxiv.org/abs/1809.05524
====================================================
SocialRobot: Towards a Personalized Elderly Care Mobile Robot (David Portugal - 14 September, 2018)
In this short paper, we make an overview of SocialRobot, the developed architecture and the human-robot interactive scenarios being prepared and tested in the framework of the project for dissemination and exploitation purposes.
Link: https://arxiv.org/abs/1809.05518
====================================================
Socially Aware Kalman Neural Networks for Trajectory Prediction (Ce Ju - 14 September, 2018)
We purpose a data-driven approach socially aware Kalman neural networks (SAKNN) where the interaction layer and the Kalman layer are embedded in the architecture, resulting in a class of architectures with huge potential to directly learn from high variance sensor input and robustly generate low variance outcomes. The evaluation of our approach on NGSIM dataset demonstrates that SAKNN performs state-of-the-art on prediction effectiveness in a relatively long-term horizon and significantly improves the signal-to-noise ratio of the predicted signal.
Link: https://arxiv.org/abs/1809.05408
====================================================
Macquarie University at BioASQ 6b: Deep learning and deep reinforcement learning for query-based multi-document summarisation (Diego Mollá - 14 September, 2018)
The deep learning architecture used features derived from the output of LSTM chains on word embeddings, plus features based on similarity with the query, and sentence position
Link: https://arxiv.org/abs/1809.05283
====================================================
Network Recasting: A Universal Method for Network Architecture Transformation (Joonsang Yu - 14 September, 2018)
Such a block-by-block recasting in a sequential manner transforms the network architecture while preserving the accuracy. It can even generate a mixed-architecture network that consists of two or more types of block
Link: https://arxiv.org/abs/1809.05262
====================================================
Do Your Cores Play Nicely? A Portable Framework for Multi-core Interference Tuning and Analysis (Dan Iorga - 13 September, 2018)
Multi-core architectures can be leveraged to allow independent processes to run in parallel. Analysing the extent of this interference is difficult due to: (1) the diversity of modern architectures, which may contain different implementations of shared resources, and (2) the complex nature of modern processors, in which interference might arise due to subtle interactions
Link: https://arxiv.org/abs/1809.05197
====================================================
IL-Net: Using Expert Knowledge to Guide the Design of Furcated Neural Networks (Khushmeen Sakloth - 13 September, 2018)
In this work, we develop a novel furcated neural network architecture that utilizes domain knowledge as high-level design principles of the network. Compared to existing state-of-the-art approaches, we show that furcated networks can improve model accuracy by approximately 20-35%, without using additional labeled data
Link: https://arxiv.org/abs/1809.05127
====================================================
Routing Cryptocurrency with the Spider Network (Vibhaalakshmi Sivaraman - 18 September, 2018)
In this paper, we present the Spider network, a new packet-switched architecture for payment channel networks. Our results show that Spider improves the volume and number of successful payments on the network by 10-45% and 5-40% respectively compared to state-of-the-art approaches.
Link: https://arxiv.org/abs/1809.05088
====================================================
DPP-PMRF: Rethinking Optimization for a Probabilistic Graphical Model Using Data-Parallel Primitives (Brenton Lessley - 13 September, 2018)
The algorithm relies on data-parallel primitives (DPPs), which provide portable performance over hardware architecture
Link: https://arxiv.org/abs/1809.05018
====================================================
Sequential Coordination of Deep Models for Learning Visual Arithmetic (Eric Crawford - 13 September, 2018)
We propose a two-tiered architecture for tackling this problem
Link: https://arxiv.org/abs/1809.04988
====================================================
Synthetic Occlusion Augmentation with Volumetric Heatmaps for the 2018 ECCV PoseTrack Challenge on 3D Human Pose Estimation (István Sárándi - 17 September, 2018)
Using a fully-convolutional backbone architecture, we obtain volumetric heatmaps per body joint, which we convert to coordinates using soft-argmax. In addition to reaching first place in the Challenge, our method also surpasses the state-of-the-art on the full Human3.6M benchmark among methods that use no additional pose datasets in training
Link: https://arxiv.org/abs/1809.04987
====================================================
High-Accuracy Inference in Neuromorphic Circuits using Hardware-Aware Training (Borna Obradovic - 13 September, 2018)
In this work, it is demonstrated that non-ideal Multi-Layer Perceptron (MLP) architectures using low bitwidth weights and activations can be trained with negligible loss of inference accuracy relative to their Floating Point-trained counterparts using a proposed off-line, continuously differentiable HW-aware training algorithm
Link: https://arxiv.org/abs/1809.04982
====================================================
Improving the Parallel Execution of Behavior Trees (Michele Colledanchise - 13 September, 2018)
The Parallel operator is rarely used despite its strong potential against other control architectures as Finite State Machines
Link: https://arxiv.org/abs/1809.04898
====================================================
Deep Network Uncertainty Maps for Indoor Navigation (Jens Lundell - 13 September, 2018)
The proposed architecture is not restricted by the assumption that the uncertainty follows a Gaussian model, as in the case of many popular solutions for deep model uncertainty estimation, e.g., MC Dropout
Link: https://arxiv.org/abs/1809.04891
====================================================
Hybrid design tools - making of a digitally augmented blackboard (Camil Octavian Milincu - 13 September, 2018)
Even though in some fields of application this approach is desirable, some particular aspects of teaching design and architecture make this transition debatable. Such a hybrid system, along with providing support for design and architecture related presentations and discussions could also mediate the contradictory relation towards technology that students and teachers have.
Link: https://arxiv.org/abs/1809.04857
====================================================
Image Captioning based on Deep Reinforcement Learning (Haichao Shi - 13 September, 2018)
However, in this paper, we propose a novel architecture for image captioning with deep reinforcement learning to optimize image captioning tasks
Link: https://arxiv.org/abs/1809.04835
====================================================
Fidelius: Protecting User Secrets from Compromised Browsers (Saba Eskandarian - 17 September, 2018)
In this work, we present Fidelius, a new architecture that uses trusted hardware enclaves integrated into the browser to enable protection of user secrets during web browsing sessions, even if the entire underlying browser and OS are fully controlled by a malicious attacker.
Link: https://arxiv.org/abs/1809.04774
====================================================
Geodesic Clustering in Deep Generative Models (Tao Yang - 12 September, 2018)
We further propose a new architecture for modeling uncertainty in variational autoencoders, which is essential for understanding the geometry of deep generative models
Link: https://arxiv.org/abs/1809.04747
====================================================
DispSegNet: Leveraging Semantics for End-to-End Learning of Disparity Estimation from Stereo Imagery (Junming Zhang - 12 September, 2018)
In this paper, we design a CNN architecture that combines these two tasks to improve the quality and accuracy of disparity estimation with the help of semantic segmentation. Experiments on KITTI and Cityscapes datasets show that our model can achieve state-of-the-art results and that leveraging embedding learned from semantic segmentation improves the performance of disparity estimation.
Link: https://arxiv.org/abs/1809.04734
====================================================
Label-less Learning for Traffic Control in an Edge Network (Min Chen - 29 August, 2018)
Specifically, we first give a statement of the problem and the system architecture
Link: https://arxiv.org/abs/1809.04525
====================================================
Multi range Real-time depth inference from a monocular stabilized footage using a Fully Convolutional Neural Network (Clément Pinard - 12 September, 2018)
Using a neural network architecture for depth map inference from monocular stabilized videos with application to UAV videos in rigid scenes, we propose a multi-range architecture for unconstrained UAV flight, leveraging flight data from sensors to make accurate depth maps for uncluttered outdoor environment. Quantitative results are given for synthetic scenes with a slightly noisy orientation, and show that our multi-range architecture improves depth inference
Link: https://arxiv.org/abs/1809.04467
====================================================
End-to-end depth from motion with stabilized monocular videos (Clément Pinard - 12 September, 2018)
We also propose an architecture for end-to-end depth inference with a fully convolutional network
Link: https://arxiv.org/abs/1809.04453
====================================================