Exclusive Style Removal for Cross Domain Novel Class Discovery

TL;DR

This paper addresses the challenge of discovering novel classes across different domains by removing style information to improve clustering, introducing a plug-in style removal module and establishing a fair benchmark for NCD.

Contribution

It proposes an exclusive style removal module for cross-domain NCD and creates a benchmark considering backbone and pre-training variations.

Findings

The style removal module improves NCD performance across datasets.

The approach is compatible with existing NCD methods as a plug-in.

Experimental results validate the effectiveness of style removal in cross-domain scenarios.

Abstract

As a promising field in open-world learning, \textit{Novel Class Discovery} (NCD) is usually a task to cluster unseen novel classes in an unlabeled set based on the prior knowledge of labeled data within the same domain. However, the performance of existing NCD methods could be severely compromised when novel classes are sampled from a different distribution with the labeled ones. In this paper, we explore and establish the solvability of NCD with cross domain setting under the necessary condition that the style information needs to be removed. Based on the theoretical analysis, we introduce an exclusive style removal module for extracting style information that is distinctive from the baseline features, thereby facilitating inference. Moreover, this module is easy to integrate with other NCD methods, acting as a plug-in to improve performance on novel classes with different distributions compared to the labeled set. Additionally, recognizing the non-negligible influence of different backbones and pre-training strategies on the performance of the NCD methods, we build a fair benchmark for future NCD research. Extensive experiments on three common datasets demonstrate the effectiveness of our proposed style removal strategy.