Fair Feature Distillation for Visual Recognition
This addresses fairness issues in computer vision systems, particularly for human-related decision-making, offering a novel approach to mitigate biases without compromising performance.
The paper tackles algorithmic bias in visual recognition by proposing MMD-based Fair Distillation (MFD), a feature distillation method that significantly reduces bias against protected groups without accuracy loss, as demonstrated on synthetic and real-world face datasets.
Fairness is becoming an increasingly crucial issue for computer vision, especially in the human-related decision systems. However, achieving algorithmic fairness, which makes a model produce indiscriminative outcomes against protected groups, is still an unresolved problem. In this paper, we devise a systematic approach which reduces algorithmic biases via feature distillation for visual recognition tasks, dubbed as MMD-based Fair Distillation (MFD). While the distillation technique has been widely used in general to improve the prediction accuracy, to the best of our knowledge, there has been no explicit work that also tries to improve fairness via distillation. Furthermore, We give a theoretical justification of our MFD on the effect of knowledge distillation and fairness. Throughout the extensive experiments, we show our MFD significantly mitigates the bias against specific minorities without any loss of the accuracy on both synthetic and real-world face datasets.