Delve into Base-Novel Confusion: Redundancy Exploration for Few-Shot Class-Incremental Learning

TL;DR

This paper investigates the confusion in few-shot class-incremental learning caused by redundancies in base-class features and proposes a method to decouple and utilize these redundancies to improve class separation and performance.

Contribution

The paper introduces Redundancy Decoupling and Integration (RDI), a novel approach that reduces confusion by managing redundancies in base-class features, enhancing FSCIL performance.

Findings

RDI effectively reduces base-novel class confusion.

The method achieves state-of-the-art results on multiple benchmarks.

Redundancy management improves feature space separation.

Abstract

Few-shot class-incremental learning (FSCIL) aims to acquire knowledge from novel classes with limited samples while retaining information about base classes. Existing methods address catastrophic forgetting and overfitting by freezing the feature extractor during novel-class learning. However, these methods usually tend to cause the confusion between base and novel classes, i.e., classifying novel-class samples into base classes. In this paper, we delve into this phenomenon to study its cause and solution. We first interpret the confusion as the collision between the novel-class and the base-class region in the feature space. Then, we find the collision is caused by the label-irrelevant redundancies within the base-class feature and pixel space. Through qualitative and quantitative experiments, we identify this redundancy as the shortcut in the base-class training, which can be decoupled to alleviate the collision. Based on this analysis, to alleviate the collision between base and novel classes, we propose a method for FSCIL named Redundancy Decoupling and Integration (RDI). RDI first decouples redundancies from base-class space to shrink the intra-base-class feature space. Then, it integrates the redundancies as a dummy class to enlarge the inter-base-class feature space. This process effectively compresses the base-class feature space, creating buffer space for novel classes and alleviating the model's confusion between the base and novel classes. Extensive experiments across benchmark datasets, including CIFAR-100, miniImageNet, and CUB-200-2011 demonstrate that our method achieves state-of-the-art performance.