Proposal-Level Unsupervised Domain Adaptation for Open World Unbiased Detector

TL;DR

This paper introduces a novel unsupervised domain adaptation approach to improve open world object detection by creating an unbiased foreground predictor that effectively distinguishes seen and unseen categories despite appearance shifts.

Contribution

It reformulates open world object detection as an unsupervised domain adaptation problem to develop an unbiased foreground predictor, enhancing detection of unseen categories.

Findings

Achieves state-of-the-art performance on OWOD benchmarks.

Effectively reduces bias towards seen categories.

Robust to appearance shifts between seen and unseen categories.

Abstract

Open World Object Detection (OWOD) combines open-set object detection with incremental learning capabilities to handle the challenge of the open and dynamic visual world. Existing works assume that a foreground predictor trained on the seen categories can be directly transferred to identify the unseen categories' locations by selecting the top-k most confident foreground predictions. However, the assumption is hardly valid in practice. This is because the predictor is inevitably biased to the known categories, and fails under the shift in the appearance of the unseen categories. In this work, we aim to build an unbiased foreground predictor by re-formulating the task under Unsupervised Domain Adaptation, where the current biased predictor helps form the domains: the seen object locations and confident background locations as the source domain, and the rest ambiguous ones as the target domain. Then, we adopt the simple and effective self-training method to learn a predictor based on the domain-invariant foreground features, hence achieving unbiased prediction robust to the shift in appearance between the seen and unseen categories. Our approach's pipeline can adapt to various detection frameworks and UDA methods, empirically validated by OWOD evaluation, where we achieve state-of-the-art performance.