SFUOD: Source-Free Unknown Object Detection

TL;DR

This paper introduces SFUOD, a novel source-free object detection framework that enables detectors to recognize both known and unknown objects in unlabeled target domains, advancing domain adaptation capabilities.

Contribution

It proposes CollaPAUL, a new method combining collaborative tuning and principal axes-based labeling to detect unknown objects without source data.

Findings

Achieves state-of-the-art results on SFUOD benchmarks.

Effectively detects unknown objects alongside known ones.

Validates robustness through extensive experiments.

Abstract

Source-free object detection adapts a detector pre-trained on a source domain to an unlabeled target domain without requiring access to labeled source data. While this setting is practical as it eliminates the need for the source dataset during domain adaptation, it operates under the restrictive assumption that only pre-defined objects from the source domain exist in the target domain. This closed-set setting prevents the detector from detecting undefined objects. To ease this assumption, we propose Source-Free Unknown Object Detection (SFUOD), a novel scenario which enables the detector to not only recognize known objects but also detect undefined objects as unknown objects. To this end, we propose CollaPAUL (Collaborative tuning and Principal Axis-based Unknown Labeling), a novel framework for SFUOD. Collaborative tuning enhances knowledge adaptation by integrating target-dependent knowledge from the auxiliary encoder with source-dependent knowledge from the pre-trained detector through a cross-domain attention mechanism. Additionally, principal axes-based unknown labeling assigns pseudo-labels to unknown objects by estimating objectness via principal axes projection and confidence scores from model predictions. The proposed CollaPAUL achieves state-of-the-art performances on SFUOD benchmarks, and extensive experiments validate its effectiveness.