CLAF: Contrastive Learning with Augmented Features for Imbalanced Semi-Supervised Learning

TL;DR

CLAF introduces a class-dependent feature augmentation method to improve contrastive semi-supervised learning on imbalanced datasets, effectively addressing minority class scarcity and bias in pseudo-labeling.

Contribution

The paper proposes CLAF, a novel contrastive learning framework with class-dependent augmentation and a strategy to select positive and negative samples from labeled data, enhancing performance on imbalanced semi-supervised tasks.

Findings

CLAF outperforms existing methods on imbalanced image classification datasets.

The class-dependent augmentation alleviates minority class sample scarcity.

Using labeled data for contrastive pairs improves robustness against class imbalance.

Abstract

Due to the advantages of leveraging unlabeled data and learning meaningful representations, semi-supervised learning and contrastive learning have been progressively combined to achieve better performances in popular applications with few labeled data and abundant unlabeled data. One common manner is assigning pseudo-labels to unlabeled samples and selecting positive and negative samples from pseudo-labeled samples to apply contrastive learning. However, the real-world data may be imbalanced, causing pseudo-labels to be biased toward the majority classes and further undermining the effectiveness of contrastive learning. To address the challenge, we propose Contrastive Learning with Augmented Features (CLAF). We design a class-dependent feature augmentation module to alleviate the scarcity of minority class samples in contrastive learning. For each pseudo-labeled sample, we select positive and negative samples from labeled data instead of unlabeled data to compute contrastive loss. Comprehensive experiments on imbalanced image classification datasets demonstrate the effectiveness of CLAF in the context of imbalanced semi-supervised learning.