KFC: Kinship Verification with Fair Contrastive Loss and Multi-Task Learning

TL;DR

This paper introduces KFC, a kinship verification model that leverages a new large dataset, multi-task learning with attention, and a fairness-aware contrastive loss to improve accuracy and reduce racial bias.

Contribution

It presents a new large dataset called KinRace, a multi-task learning framework with attention, and a novel fairness-aware contrastive loss for kinship verification.

Findings

Surpasses state-of-the-art accuracy

Effectively reduces racial bias in kinship verification

Demonstrates robustness across diverse datasets

Abstract

Kinship verification is an emerging task in computer vision with multiple potential applications. However, there's no large enough kinship dataset to train a representative and robust model, which is a limitation for achieving better performance. Moreover, face verification is known to exhibit bias, which has not been dealt with by previous kinship verification works and sometimes even results in serious issues. So we first combine existing kinship datasets and label each identity with the correct race in order to take race information into consideration and provide a larger and complete dataset, called KinRace dataset. Secondly, we propose a multi-task learning model structure with attention module to enhance accuracy, which surpasses state-of-the-art performance. Lastly, our fairness-aware contrastive loss function with adversarial learning greatly mitigates racial bias. We introduce a debias term into traditional contrastive loss and implement gradient reverse in race classification task, which is an innovative idea to mix two fairness methods to alleviate bias. Exhaustive experimental evaluation demonstrates the effectiveness and superior performance of the proposed KFC in both standard deviation and accuracy at the same time.