Cross-domain Open-world Discovery

TL;DR

This paper introduces CROW, a prototype-based method for cross-domain open-world discovery that effectively identifies novel classes under domain shifts, outperforming existing methods on image classification benchmarks.

Contribution

CROW is a novel approach that combines clustering and matching in a structured representation space to discover unseen classes across domains.

Findings

CROW achieves an 8% average performance improvement over baselines.

It effectively discovers novel classes under domain shifts.

Extensive experiments validate its superior performance.

Abstract

In many real-world applications, test data may commonly exhibit categorical shifts, characterized by the emergence of novel classes, as well as distribution shifts arising from feature distributions different from the ones the model was trained on. However, existing methods either discover novel classes in the open-world setting or assume domain shifts without the ability to discover novel classes. In this work, we consider a cross-domain open-world discovery setting, where the goal is to assign samples to seen classes and discover unseen classes under a domain shift. To address this challenging problem, we present CROW, a prototype-based approach that introduces a cluster-then-match strategy enabled by a well-structured representation space of foundation models. In this way, CROW discovers novel classes by robustly matching clusters with previously seen classes, followed by fine-tuning the representation space using an objective designed for cross-domain open-world discovery. Extensive experimental results on image classification benchmark datasets demonstrate that CROW outperforms alternative baselines, achieving an 8% average performance improvement across 75 experimental settings.