Unbiased Model Prediction Without Using Protected Attribute Information

TL;DR

This paper introduces NPAD, a novel bias mitigation algorithm that does not require protected attribute data, using non-protected attributes and new loss functions to reduce bias in facial attribute prediction.

Contribution

The paper proposes NPAD, a bias mitigation method that avoids using protected attributes, with two new loss functions, demonstrating effectiveness on facial datasets.

Findings

Significant bias reduction across gender and age groups.

Effective bias mitigation without protected attribute data.

Proposed loss functions improve fairness in facial attribute prediction.

Abstract

The problem of bias persists in the deep learning community as models continue to provide disparate performance across different demographic subgroups. Therefore, several algorithms have been proposed to improve the fairness of deep models. However, a majority of these algorithms utilize the protected attribute information for bias mitigation, which severely limits their application in real-world scenarios. To address this concern, we have proposed a novel algorithm, termed as \textbf{Non-Protected Attribute-based Debiasing (NPAD)} algorithm for bias mitigation, that does not require the protected attribute information. The proposed NPAD algorithm utilizes the auxiliary information provided by the non-protected attributes to optimize the model for bias mitigation. Further, two different loss functions, \textbf{Debiasing via Attribute Cluster Loss (DACL)} and \textbf{Filter Redundancy Loss (FRL)} have been proposed to optimize the model for fairness goals. Multiple experiments are performed on the LFWA and CelebA datasets for facial attribute prediction, and a significant reduction in bias across different gender and age subgroups is observed.