A class-incremental learning approach for learning feature-compatible embeddings

doi:10.1016/j.neunet.2024.106685

2024-12-01

PRE

DOI：10.1016/j.neunet.2024.106685

原文链接

原文求助

摘要

En 中文

Humans have the ability to constantly learn new knowledge. However, for artificial intelligence, trying to continuously learn new knowledge usually results in catastrophic forgetting, the existing regularization- based and dynamic structure-based approaches have shown great potential for alleviating. Nevertheless, these approaches have certain limitations. They usually do not fully consider the problem of incompatible feature embeddings. Instead, they tend to focus only on the features of new or previous classes and fail to comprehensively consider the entire model. Therefore, we propose a two-stage learning paradigm to solve feature embedding incompatibility problems. Specifically, we retain the previous model and freeze all its parameters in the first stage while dynamically expanding a new module to alleviate feature embedding incompatibility questions. In the second stage, a fusion knowledge distillation approach is used to compress the redundant feature dimensions. Moreover, we propose weight pruning and consolidation approaches to improve the efficiency of the model. Our experimental results obtained on the CIFAR-100, ImageNet-100 and ImageNet-1000 benchmark datasets show that the proposed approaches achieve the best performance among all the compared approaches. For example, on the ImageNet-100 dataset, the maximal accuracy improvement is 5.08%. Code is available at https://github.com/ybyangjing/CIL-FCE.

Keyword:

Class-incremental learning

Catastrophic forgetting

Feature embedding

Knowledge distillation