Fine-Grained Knowledge Selection and Restoration for Non-Exemplar Class Incremental Learning

TL;DR

This paper introduces a novel framework for non-exemplar class incremental learning that uses fine-grained knowledge selection and restoration to better balance learning new tasks while preventing forgetting, without access to previous data.

Contribution

It proposes a fine-grained selective patch-level distillation method and a task-agnostic prototype generation mechanism to improve incremental learning performance without past data.

Findings

Effective on CIFAR100, TinyImageNet, ImageNet-Subset

Outperforms existing methods in preventing catastrophic forgetting

Enhances knowledge retention and new task learning balance

Abstract

Non-exemplar class incremental learning aims to learn both the new and old tasks without accessing any training data from the past. This strict restriction enlarges the difficulty of alleviating catastrophic forgetting since all techniques can only be applied to current task data. Considering this challenge, we propose a novel framework of fine-grained knowledge selection and restoration. The conventional knowledge distillation-based methods place too strict constraints on the network parameters and features to prevent forgetting, which limits the training of new tasks. To loose this constraint, we proposed a novel fine-grained selective patch-level distillation to adaptively balance plasticity and stability. Some task-agnostic patches can be used to preserve the decision boundary of the old task. While some patches containing the important foreground are favorable for learning the new task. Moreover, we employ a task-agnostic mechanism to generate more realistic prototypes of old tasks with the current task sample for reducing classifier bias for fine-grained knowledge restoration. Extensive experiments on CIFAR100, TinyImageNet and ImageNet-Subset demonstrate the effectiveness of our method. Code is available at https://github.com/scok30/vit-cil.