Continual Contrastive Learning for Image Classification

TL;DR

This paper introduces a continual contrastive learning framework that addresses catastrophic forgetting in streaming data, improving image classification accuracy over existing self-supervised methods.

Contribution

It proposes a rehearsal-based framework with a novel sampling strategy, knowledge distillation, and an extra sample queue to enhance continual learning in contrastive self-supervised methods.

Findings

Improves CIFAR-100 accuracy by 1.60%

Enhances ImageNet-Sub accuracy by 2.86%

Boosts ImageNet-Full accuracy by 1.29%

Abstract

Recently, self-supervised representation learning gives further development in multimedia technology. Most existing self-supervised learning methods are applicable to packaged data. However, when it comes to streamed data, they are suffering from a catastrophic forgetting problem, which is not studied extensively. In this paper, we make the first attempt to tackle the catastrophic forgetting problem in the mainstream self-supervised methods, i.e., contrastive learning methods. Specifically, we first develop a rehearsal-based framework combined with a novel sampling strategy and a self-supervised knowledge distillation to transfer information over time efficiently. Then, we propose an extra sample queue to help the network separate the feature representations of old and new data in the embedding space. Experimental results show that compared with the naive self-supervised baseline, which learns tasks one by one without taking any technique, we improve the image classification accuracy by 1.60% on CIFAR-100, 2.86% on ImageNet-Sub, and 1.29% on ImageNet-Full under 10 incremental steps setting. Our code will be available at https://github.com/VDIGPKU/ContinualContrastiveLearning.