IMEX-Reg: Implicit-Explicit Regularization in the Function Space for Continual Learning

TL;DR

IMEX-Reg introduces a novel implicit-explicit regularization method leveraging contrastive learning and consistency regularization to enhance continual learning, especially under low-buffer regimes, by reducing catastrophic forgetting and improving generalization.

Contribution

The paper proposes IMEX-Reg, a new regularization approach combining contrastive learning and consistency regularization, with theoretical support, to improve continual learning performance in low-buffer scenarios.

Findings

IMEX-Reg outperforms existing rehearsal methods in various CL scenarios.

It demonstrates robustness to natural and adversarial corruptions.

The approach reduces task-recency bias and improves generalization.

Abstract

Continual learning (CL) remains one of the long-standing challenges for deep neural networks due to catastrophic forgetting of previously acquired knowledge. Although rehearsal-based approaches have been fairly successful in mitigating catastrophic forgetting, they suffer from overfitting on buffered samples and prior information loss, hindering generalization under low-buffer regimes. Inspired by how humans learn using strong inductive biases, we propose IMEX-Reg to improve the generalization performance of experience rehearsal in CL under low buffer regimes. Specifically, we employ a two-pronged implicit-explicit regularization approach using contrastive representation learning (CRL) and consistency regularization. To further leverage the global relationship between representations learned using CRL, we propose a regularization strategy to guide the classifier toward the activation correlations in the unit hypersphere of the CRL. Our results show that IMEX-Reg significantly improves generalization performance and outperforms rehearsal-based approaches in several CL scenarios. It is also robust to natural and adversarial corruptions with less task-recency bias. Additionally, we provide theoretical insights to support our design decisions further.