ECL: Class-Enhancement Contrastive Learning for Long-tailed Skin Lesion Classification

TL;DR

This paper introduces ECL, a contrastive learning method that enhances minority class information and balances class treatment, significantly improving long-tailed skin lesion classification performance.

Contribution

The paper proposes a novel class-Enhancement Contrastive Learning approach with hybrid-proxy models and balanced loss functions to address class imbalance in skin lesion datasets.

Findings

ECL outperforms existing methods on imbalanced skin lesion datasets.

The hybrid-proxy model effectively generates class-dependent proxies.

Balanced loss functions improve minority class recognition.

Abstract

Skin image datasets often suffer from imbalanced data distribution, exacerbating the difficulty of computer-aided skin disease diagnosis. Some recent works exploit supervised contrastive learning (SCL) for this long-tailed challenge. Despite achieving significant performance, these SCL-based methods focus more on head classes, yet ignoring the utilization of information in tail classes. In this paper, we propose class-Enhancement Contrastive Learning (ECL), which enriches the information of minority classes and treats different classes equally. For information enhancement, we design a hybrid-proxy model to generate class-dependent proxies and propose a cycle update strategy for parameters optimization. A balanced-hybrid-proxy loss is designed to exploit relations between samples and proxies with different classes treated equally. Taking both "imbalanced data" and "imbalanced diagnosis difficulty" into account, we further present a balanced-weighted cross-entropy loss following curriculum learning schedule. Experimental results on the classification of imbalanced skin lesion data have demonstrated the superiority and effectiveness of our method.