Knowledge Restore and Transfer for Multi-label Class-Incremental Learning

TL;DR

This paper introduces a novel framework for multi-label class-incremental learning that effectively restores and transfers knowledge, addressing challenges like label absence and information dilution, and demonstrates improved performance on benchmark datasets.

Contribution

The paper proposes a knowledge restore and transfer framework with dynamic pseudo-labels and incremental cross-attention modules for multi-label incremental learning, a less-studied area.

Findings

Improves recognition performance on MS-COCO and PASCAL VOC datasets.

Effectively mitigates catastrophic forgetting in multi-label incremental learning.

Demonstrates superiority over existing methods in experimental results.

Abstract

Current class-incremental learning research mainly focuses on single-label classification tasks while multi-label class-incremental learning (MLCIL) with more practical application scenarios is rarely studied. Although there have been many anti-forgetting methods to solve the problem of catastrophic forgetting in class-incremental learning, these methods have difficulty in solving the MLCIL problem due to label absence and information dilution. In this paper, we propose a knowledge restore and transfer (KRT) framework for MLCIL, which includes a dynamic pseudo-label (DPL) module to restore the old class knowledge and an incremental cross-attention(ICA) module to save session-specific knowledge and transfer old class knowledge to the new model sufficiently. Besides, we propose a token loss to jointly optimize the incremental cross-attention module. Experimental results on MS-COCO and PASCAL VOC datasets demonstrate the effectiveness of our method for improving recognition performance and mitigating forgetting on multi-label class-incremental learning tasks.