Continual Novelty Detection

TL;DR

This paper explores the integration of novelty detection into continual learning, highlighting its importance for flexible, transparent AI systems and providing benchmarks and insights into their interaction.

Contribution

It formulates the problem of continual novelty detection, introduces a benchmark for evaluation, and discusses the impact of continual learning on novelty detection methods.

Findings

Continual learning influences novelty detection behavior.

Novelty detection can reveal insights about continual learners.

Benchmark results show varying performance of methods under different settings.

Abstract

Novelty Detection methods identify samples that are not representative of a model's training set thereby flagging misleading predictions and bringing a greater flexibility and transparency at deployment time. However, research in this area has only considered Novelty Detection in the offline setting. Recently, there has been a growing realization in the computer vision community that applications demand a more flexible framework - Continual Learning - where new batches of data representing new domains, new classes or new tasks become available at different points in time. In this setting, Novelty Detection becomes more important, interesting and challenging. This work identifies the crucial link between the two problems and investigates the Novelty Detection problem under the Continual Learning setting. We formulate the Continual Novelty Detection problem and present a benchmark, where we compare several Novelty Detection methods under different Continual Learning settings. We show that Continual Learning affects the behaviour of novelty detection algorithms, while novelty detection can pinpoint insights in the behaviour of a continual learner. We further propose baselines and discuss possible research directions. We believe that the coupling of the two problems is a promising direction to bring vision models into practice.