Enhancing Fairness of Visual Attribute Predictors
It addresses fairness issues in image recognition for underrepresented groups, though it appears incremental as it builds on existing fairness metrics.
The paper tackles bias in deep neural networks for visual attribute prediction by introducing fairness-aware regularization losses, showing effectiveness in improving fairness while maintaining high classification performance on datasets like CelebA, UTKFace, and SIIM-ISIC.
The performance of deep neural networks for image recognition tasks such as predicting a smiling face is known to degrade with under-represented classes of sensitive attributes. We address this problem by introducing fairness-aware regularization losses based on batch estimates of Demographic Parity, Equalized Odds, and a novel Intersection-over-Union measure. The experiments performed on facial and medical images from CelebA, UTKFace, and the SIIM-ISIC melanoma classification challenge show the effectiveness of our proposed fairness losses for bias mitigation as they improve model fairness while maintaining high classification performance. To the best of our knowledge, our work is the first attempt to incorporate these types of losses in an end-to-end training scheme for mitigating biases of visual attribute predictors. Our code is available at https://github.com/nish03/FVAP.