netopedro/Continuous-Ethnicity-Face-Recognition

Dataset Card for Ethnicity Fairness in a Continuous Space

Dataset Details

Dataset Description

This dataset provides the training images (from BalancedFace and GlobalFace produced by BUPT) used in the paper: "Balancing Beyond Discrete Categories: Continuous Demographic Labels for Fair Face Recognition".

These have been curated to be balanced in a continuous ethnicity space, following three different strategies: Protocol A, Protocol B and Protocol C.

In addition to these protocols, one can include model-based fairness approaches.

• Curated by: Pedro C. Neto, Naser Damer, Jaime S. Cardoso, Ana F. Sequeira

Dataset Sources [optional]

• Repository: GitHub
• Paper: Arxiv

Uses

Direct Use

This dataset can be used to train fair face recognition methods. Using the provided list of images, one can quickly read the corresponding image on BUPT datasets.

Dataset Structure

[More Information Needed]

Dataset Creation

Source Data

BUPT-BalancedFace and BUPT-GlobalFace

Citation

BibTeX:

@article{neto2025balancing,
  title={Balancing Beyond Discrete Categories: Continuous Demographic Labels for Fair Face Recognition},
  author={Neto, Pedro C and Damer, Naser and Cardoso, Jaime S and Sequeira, Ana F},
  journal={arXiv preprint arXiv:2506.01532},
  year={2025}
}