metadata
license: cc-by-nc-nd-4.0
task_categories:
- image-classification
- image-segmentation
tags:
- ear
- Machine learning
- Classification
- Forensics
- Detection
size_categories:
- 10K<n<100K
Ear Detection - 14,000+ Images
The dataset comprises 14,000+ ear images from 2,000 unique individuals, paired with reference face photos and demographic labels. Designed for ear recognition and biometric identification, it helps research in human ear detection, recognition accuracy, and biometric systems.— Get the data
Dataset characteristics:
| Characteristic | Data |
|---|---|
| Description | Images of an ear with a face photo |
| Data types | Image |
| Tasks | Ear‑biometric R&D, Forensic comparison, and missing‑person identification |
| Number of images | 14,000 |
| Number of people | 2,000 |
| Number of files in a set | 7 (Full face - 1, Left ear - 3, Right ear - 3) |
| Labeling | Metadata (age, gender, ethnicity, side, proximity, lighting) |
| Gender | Male, Female |
📊 Sample dataset available! For full access, contact us to discuss purchase terms.
Dataset structure
- Black_Africandescent — a folder of photos of people from this ethnicity.
- Caucasian — a folder of photos of people from this ethnicity.
- East_SouthEast Asian — a folder of photos of people from this ethnicity.
- Latinx_Hispanic — a folder of photos of people from this ethnicity.
- South Asian — a folder of photos of people from this ethnicity.
- Meta.csv — file contains metadata for all individuals in the dataset.