Dynamic Feature Learning and Matching for Class-Incremental Learning

TL;DR

This paper introduces a novel dynamic feature learning and matching approach for class-incremental learning, addressing data augmentation, feature discriminativeness, and classifier alignment to improve performance.

Contribution

It proposes the DFLM model that integrates class weight info, non-stationary functions, vMF classifier, and matching loss for enhanced dynamic feature learning.

Findings

Significant performance improvements on CIL benchmarks.

Effective modeling of dynamic feature distributions.

Improved alignment between features and classifier.

Abstract

Class-incremental learning (CIL) has emerged as a means to learn new classes incrementally without catastrophic forgetting of previous classes. Recently, CIL has undergone a paradigm shift towards dynamic architectures due to their superior performance. However, these models are still limited by the following aspects: (i) Data augmentation (DA), which are tightly coupled with CIL, remains under-explored in dynamic architecture scenarios. (ii) Feature representation. The discriminativeness of dynamic feature are sub-optimal and possess potential for refinement. (iii) Classifier. The misalignment between dynamic feature and classifier constrains the capabilities of the model. To tackle the aforementioned drawbacks, we propose the Dynamic Feature Learning and Matching (DFLM) model in this paper from above three perspectives. Specifically, we firstly introduce class weight information and non-stationary functions to extend the mix DA method for dynamically adjusting the focus on memory during training. Then, von Mises-Fisher (vMF) classifier is employed to effectively model the dynamic feature distribution and implicitly learn their discriminative properties. Finally, the matching loss is proposed to facilitate the alignment between the learned dynamic features and the classifier by minimizing the distribution distance. Extensive experiments on CIL benchmarks validate that our proposed model achieves significant performance improvements over existing methods.