ChiMera: Learning with noisy labels by contrasting mixed-up augmentations

TL;DR

ChiMera introduces a novel contrastive learning approach called MixCLR within a semi-supervised framework to effectively learn from noisy labels, achieving state-of-the-art results on multiple real-world datasets.

Contribution

The paper proposes MixCLR, a new contrastive learning technique for noisy labels, integrated into ChiMera, a semi-supervised framework that enhances label noise robustness.

Findings

Achieves state-of-the-art performance on seven datasets.

Effectively handles both symmetric and asymmetric label noise.

Improves the robustness of learned representations against label noise.

Abstract

Learning with noisy labels has been studied to address incorrect label annotations in real-world applications. In this paper, we present ChiMera, a two-stage learning-from-noisy-labels framework based on semi-supervised learning, developed based on a novel contrastive learning technique MixCLR. The key idea of MixCLR is to learn and refine the representations of mixed augmentations from two different images to better resist label noise. ChiMera jointly learns the representations of the original data distribution and mixed-up data distribution via MixCLR, introducing many additional augmented samples to fill in the gap between different classes. This results in a more smoothed representation space learned by contrastive learning with better alignment and a more robust decision boundary. By exploiting MixCLR, ChiMera also improves the label diffusion process in the semi-supervised noise recovery stage and further boosts its ability to diffuse correct label information. We evaluated ChiMera on seven real-world datasets and obtained state-of-the-art performance on both symmetric noise and asymmetric noise. Our method opens up new avenues for using contrastive learning on learning with noisy labels and we envision MixCLR to be broadly applicable to other applications.