Motivation
TAMEn provides an excellent closed-loop tactile data collection pipeline for contact-rich tasks. As the tactile sensing ecosystem grows, researchers often work with multiple sensor types (GelSight, DIGIT, PaXini, etc.) and need a standardized way to exchange and compare data across platforms.
Proposal
Consider supporting TLabel as an optional export format. TLabel provides:
- Unified data schema (
tlabel.core.TLabelData) for structured tactile data with sensor metadata
- 10+ sensor adapters out of the box (GelSight, PaXini, Daimon, UniVTAC, VTouch, YCB-Slide, TacQuad, etc.)
- CLI tools for format conversion, dataset management, and quality checks
- Lightweight:
pip install tlabel, MIT licensed
This would enable TAMEn users to:
- Export collected data in a standardized format compatible with other tactile datasets
- Easily compare results across different sensor platforms
- Leverage TLabel's growing ecosystem of tools and converters
Implementation
A simple exporter could convert TAMEn's internal data format to TLabel's schema. Happy to help with implementation or testing if the maintainers are interested.
References
Motivation
TAMEn provides an excellent closed-loop tactile data collection pipeline for contact-rich tasks. As the tactile sensing ecosystem grows, researchers often work with multiple sensor types (GelSight, DIGIT, PaXini, etc.) and need a standardized way to exchange and compare data across platforms.
Proposal
Consider supporting TLabel as an optional export format. TLabel provides:
tlabel.core.TLabelData) for structured tactile data with sensor metadatapip install tlabel, MIT licensedThis would enable TAMEn users to:
Implementation
A simple exporter could convert TAMEn's internal data format to TLabel's schema. Happy to help with implementation or testing if the maintainers are interested.
References