Let Samples Speak: Mitigating Spurious Correlation by Exploiting the Clusterness of Samples

TL;DR

This paper introduces a data-driven method to reduce spurious correlations in deep learning by exploiting sample clustering, leading to more unbiased models with significantly improved worst group accuracy.

Contribution

It proposes a novel pipeline that identifies, neutralizes, and eliminates spurious features through sample clustering and feature transformation, enhancing model robustness.

Findings

Over 20% improvement in worst group accuracy on benchmarks

Effective in both image and NLP debiasing tasks

Outperforms standard ERM methods significantly

Abstract

Deep learning models are known to often learn features that spuriously correlate with the class label during training but are irrelevant to the prediction task. Existing methods typically address this issue by annotating potential spurious attributes, or filtering spurious features based on some empirical assumptions (e.g., simplicity of bias). However, these methods may yield unsatisfactory performance due to the intricate and elusive nature of spurious correlations in real-world data. In this paper, we propose a data-oriented approach to mitigate the spurious correlation in deep learning models. We observe that samples that are influenced by spurious features tend to exhibit a dispersed distribution in the learned feature space. This allows us to identify the presence of spurious features. Subsequently, we obtain a bias-invariant representation by neutralizing the spurious features based on a simple grouping strategy. Then, we learn a feature transformation to eliminate the spurious features by aligning with this bias-invariant representation. Finally, we update the classifier by incorporating the learned feature transformation and obtain an unbiased model. By integrating the aforementioned identifying, neutralizing, eliminating and updating procedures, we build an effective pipeline for mitigating spurious correlation. Experiments on image and NLP debiasing benchmarks show an improvement in worst group accuracy of more than 20% compared to standard empirical risk minimization (ERM). Codes and checkpoints are available at https://github.com/davelee-uestc/nsf_debiasing .