Age Estimation Face Dataset: Minors & Young Adults (10-30 Years)

Age estimation face dataset: 10,000+ consented selfies of minors & young adults (10-30 years) with verified per-year age labels. Multi-ethnic, phone-captured. Built for under-18 age gating, age verification, and face recognition.

Dataset Overview

Total images	10,000+
Age range	10-30 (per-year folders)
Ethnicities	Black, Caucasian, Hispanic
Gender	Male, Female
Capture device	Smartphone front camera
Labels	age, gender, ethnicity
Format	JPEG images + CSV metadata
Consent	Full legal consent

Metadata Format

Column	Description	Example
file_name	Image filename	age_14_0023.jpeg
age_folder	Age folder	age_14
age	Verified age at capture	14
gender	Gender	female
ethnicity	Ethnicity	black

Use Cases

Age Gating & Legal Compliance

Input: User selfie at app sign-up
Model: Age estimation (trained on this dataset)
Output: under_18 / over_18
Action: Block or allow access

Train binary classifiers or regression models that accurately detect whether a user is under 18. Dense per-year sampling around the 18-year boundary reduces false positives and false negatives where it matters most.

Age Estimation

Fine-tune pretrained models (DEX, MiVOLO, AgeNet) on the 10-30 segment to reduce MAE in the range where public datasets underperform. Verified birth-year labels provide clean ground truth.

Age Verification Without ID

Build selfie-only age checks for platforms that need frictionless onboarding. No document upload required — the model estimates age directly from a phone-captured photo.

Bias Evaluation

Test your existing age estimation model across ethnicities and genders using the structured metadata. Identify and fix demographic accuracy gaps before production deployment.

Comparison With Public Datasets

Feature	This Dataset	UTKFace	IMDB-WIKI	MORPH	FG-NET
Age range	10-30	0-116	0-100	16-77	0-69
Focus on teens	Dense	Sparse	Sparse	None	Sparse
Label quality	Birth-year verified	Filename-parsed	Noisy/scraped	Verified	Verified
Ethnic diversity	3 groups, balanced	5 groups	Skewed	80% one group	Limited
Consent	Full + guardian	None	None	Research only	Research only
Commercial license	Yes	No	No	No	No
Selfie-specific	Yes	Mixed	Web photos	Mugshot-style	Mixed

Getting the Full Dataset

This repository contains a descriprion only. For the samplee and full 10,000+ image dataset with commercial license:

1. Browse samples on Kaggle or HuggingFace
2. Request access at axonlab.ai
3. Contact us at sales@axonlabs.pro