Detachedly Learn a Classifier for Class-Incremental Learning

TL;DR

This paper introduces a task-aware experience replay strategy for class-incremental learning that preserves previous knowledge and reduces overfitting, leading to improved classifier performance.

Contribution

It proposes a novel replay method that rebalances loss and detaches old classifier weights, addressing knowledge degradation in continual learning.

Findings

Outperforms current state-of-the-art methods

Effectively preserves previous knowledge

Reduces overfitting on episodic memory

Abstract

In continual learning, model needs to continually learn a feature extractor and classifier on a sequence of tasks. This paper focuses on how to learn a classifier based on a pretrained feature extractor under continual learning setting. We present an probabilistic analysis that the failure of vanilla experience replay (ER) comes from unnecessary re-learning of previous tasks and incompetence to distinguish current task from the previous ones, which is the cause of knowledge degradation and prediction bias. To overcome these weaknesses, we propose a novel replay strategy task-aware experience replay. It rebalances the replay loss and detaches classifier weight for the old tasks from the update process, by which the previous knowledge is kept intact and the overfitting on episodic memory is alleviated. Experimental results show our method outperforms current state-of-the-art methods.