Continual Contrastive Learning for Image Classification
This addresses the issue of forgetting in continual learning for AI systems handling sequential data, representing an incremental advance in adapting self-supervised methods to dynamic environments.
The paper tackles the catastrophic forgetting problem in self-supervised contrastive learning for streamed data by proposing a rehearsal-based framework with novel sampling and knowledge distillation, improving image classification accuracy by up to 2.86% on benchmarks like CIFAR-100 and ImageNet.
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.