jiazhi412 / Representation-Level-Bias

Representation Level Bias (RLB) Assessment

Information-Theoretic Bias Assessment Of Learned Representations Of Pretrained Face Recognition

Jiazhi Li, Wael Abd-Almageed

@inproceedings{li2021information,
  title={Information-Theoretic Bias Assessment Of Learned Representations Of Pretrained Face Recognition},
  author={Li, Jiazhi and Abd-Almageed, Wael},
  booktitle={2021 16th IEEE International Conference on Automatic Face and Gesture Recognition (FG 2021)},
  pages={1--8},
  year={2021},
  organization={IEEE}
}

Requirements

pip install -r requirements.txt

Experiments

To conduct experiments, run

bash run_colored_model.sh

After running, the convergence lines shown as Fig.3(a) in the paper will be saved in the current directory as colored_mnist_convergence.png. The results with each different color standard deviation will be saved in ./result/colored_model/reproduce/.