Component-Based Fairness in Face Attribute Classification with Bayesian Network-informed Meta Learning

Code implementation for FAccT 2025 paper "Component-Based Fairness in Face Attribute Classification with Bayesian Network-informed Meta Learning"

Requirement

To prepare the environment, use ./environment.txt file and conda.

Training

The training entry is in ./src/train.sh. In train.sh, replace the data_root argument with the path to folder containing CelebA dataset.

Visualizations and Models

Please find the learned Bayesian Network under BNMR/save/(test)bnn[1, 6, 7, 14, 24].xml, accompanied with the visualization for 5 attributes and all attributes DAG.

Acknowledgement

During the implementation we base our code mostly on Transformers from HuggingFace and Meta-Weight-Net Many thanks to the authors for their great work!

Cite

Please consider citing the following papers if you use our methods/code in your research:

@inproceedings{10.1145/3715275.3732066,
author = {Liu, Yifan and Yao, Ruichen and Liu, Yaokun and Zong, Ruohan and Li, Zelin and Zhang, Yang and Wang, Dong},
title = {Component-Based Fairness in Face Attribute Classification with Bayesian Network-informed Meta Learning},
year = {2025},
isbn = {9798400714825},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
url = {https://doi.org/10.1145/3715275.3732066},
doi = {10.1145/3715275.3732066},
pages = {1015–1027},
numpages = {13},
keywords = {Fairness, Face Attribute Classification, Bayesian Network, Meta Learning, Sample Reweighting},
location = {
},
series = {FAccT '25}
}