DSS: A Diverse Sample Selection Method to Preserve Knowledge in Class-Incremental Learning

TL;DR

This paper introduces DSS, a simple yet effective method for selecting diverse samples in class-incremental learning, which improves knowledge retention by better exemplar diversity under various task boundary scenarios.

Contribution

DSS is a novel sample selection technique that enhances exemplar diversity in class-incremental learning, outperforming existing methods with simpler implementation.

Findings

DSS outperforms state-of-the-art methods in class-incremental learning.

DSS is effective under both disjoint and fuzzy task boundary scenarios.

DSS is simpler to understand and implement.

Abstract

Rehearsal-based techniques are commonly used to mitigate catastrophic forgetting (CF) in Incremental learning (IL). The quality of the exemplars selected is important for this purpose and most methods do not ensure the appropriate diversity of the selected exemplars. We propose a new technique "DSS" -- Diverse Selection of Samples from the input data stream in the Class-incremental learning (CIL) setup under both disjoint and fuzzy task boundary scenarios. Our method outperforms state-of-the-art methods and is much simpler to understand and implement.