Language-guided Detection and Mitigation of Unknown Dataset Bias

TL;DR

This paper introduces a framework for detecting and mitigating dataset bias in classifiers without prior bias knowledge, using caption analysis and data augmentation, leading to improved fairness and performance.

Contribution

It presents a novel bias detection method based on caption keywords and two debiasing techniques, outperforming prior approaches that require prior bias knowledge.

Findings

Outperforms existing methods without prior bias knowledge

Achieves comparable results to methods with prior bias information

Enhances classifier fairness and accuracy on biased datasets

Abstract

Dataset bias is a significant problem in training fair classifiers. When attributes unrelated to classification exhibit strong biases towards certain classes, classifiers trained on such dataset may overfit to these bias attributes, substantially reducing the accuracy for minority groups. Mitigation techniques can be categorized according to the availability of bias information (\ie, prior knowledge). Although scenarios with unknown biases are better suited for real-world settings, previous work in this field often suffers from a lack of interpretability regarding biases and lower performance. In this study, we propose a framework to identify potential biases as keywords without prior knowledge based on the partial occurrence in the captions. We further propose two debiasing methods: (a) handing over to an existing debiasing approach which requires prior knowledge by assigning pseudo-labels, and (b) employing data augmentation via text-to-image generative models, using acquired bias keywords as prompts. Despite its simplicity, experimental results show that our framework not only outperforms existing methods without prior knowledge, but also is even comparable with a method that assumes prior knowledge.