Improving Multi-Label Contrastive Learning by Leveraging Label Distribution

TL;DR

This paper introduces a novel multi-label contrastive learning approach that leverages label distribution to improve sample selection and label relationship modeling, leading to better representations.

Contribution

It proposes a new method that simplifies positive/negative sample selection and models label relationships using label distributions derived from logical labels.

Findings

Outperforms state-of-the-art methods on six metrics

Effective label distribution modeling improves contrastive learning

Applicable to diverse multi-label datasets

Abstract

In multi-label learning, leveraging contrastive learning to learn better representations faces a key challenge: selecting positive and negative samples and effectively utilizing label information. Previous studies selected positive and negative samples based on the overlap between labels and used them for label-wise loss balancing. However, these methods suffer from a complex selection process and fail to account for the varying importance of different labels. To address these problems, we propose a novel method that improves multi-label contrastive learning through label distribution. Specifically, when selecting positive and negative samples, we only need to consider whether there is an intersection between labels. To model the relationships between labels, we introduce two methods to recover label distributions from logical labels, based on Radial Basis Function (RBF) and contrastive loss, respectively. We evaluate our method on nine widely used multi-label datasets, including image and vector datasets. The results demonstrate that our method outperforms state-of-the-art methods in six evaluation metrics.