Does Data Repair Lead to Fair Models? Curating Contextually Fair Data To Reduce Model Bias

TL;DR

This paper introduces a data repair algorithm that curates fair, balanced datasets by reducing co-occurrence bias, leading to fairer model predictions across various tasks without sacrificing overall accuracy.

Contribution

The paper presents a novel, simple data repair method using the coefficient of variation to create contextually fair datasets for training unbiased models across different tasks.

Findings

Curating fair data improves model fairness across protected groups.

The method maintains overall model performance while reducing bias.

Applicable to various tasks and training scenarios.

Abstract

Contextual information is a valuable cue for Deep Neural Networks (DNNs) to learn better representations and improve accuracy. However, co-occurrence bias in the training dataset may hamper a DNN model's generalizability to unseen scenarios in the real world. For example, in COCO, many object categories have a much higher co-occurrence with men compared to women, which can bias a DNN's prediction in favor of men. Recent works have focused on task-specific training strategies to handle bias in such scenarios, but fixing the available data is often ignored. In this paper, we propose a novel and more generic solution to address the contextual bias in the datasets by selecting a subset of the samples, which is fair in terms of the co-occurrence with various classes for a protected attribute. We introduce a data repair algorithm using the coefficient of variation, which can curate fair and contextually balanced data for a protected class(es). This helps in training a fair model irrespective of the task, architecture or training methodology. Our proposed solution is simple, effective, and can even be used in an active learning setting where the data labels are not present or being generated incrementally. We demonstrate the effectiveness of our algorithm for the task of object detection and multi-label image classification across different datasets. Through a series of experiments, we validate that curating contextually fair data helps make model predictions fair by balancing the true positive rate for the protected class across groups without compromising on the model's overall performance.