Learning Domain Adaptive Object Detection with Probabilistic Teacher

TL;DR

This paper introduces Probabilistic Teacher, a framework for unsupervised domain adaptive object detection that models uncertainty in pseudo labels to improve training without relying on confidence thresholds.

Contribution

It proposes a novel uncertainty-guided self-training framework with entropy focal loss, enhancing domain adaptation performance in object detection.

Findings

Outperforms previous methods significantly

Achieves new state-of-the-art results

Effectively models pseudo label uncertainty

Abstract

Self-training for unsupervised domain adaptive object detection is a challenging task, of which the performance depends heavily on the quality of pseudo boxes. Despite the promising results, prior works have largely overlooked the uncertainty of pseudo boxes during self-training. In this paper, we present a simple yet effective framework, termed as Probabilistic Teacher (PT), which aims to capture the uncertainty of unlabeled target data from a gradually evolving teacher and guides the learning of a student in a mutually beneficial manner. Specifically, we propose to leverage the uncertainty-guided consistency training to promote classification adaptation and localization adaptation, rather than filtering pseudo boxes via an elaborate confidence threshold. In addition, we conduct anchor adaptation in parallel with localization adaptation, since anchor can be regarded as a learnable parameter. Together with this framework, we also present a novel Entropy Focal Loss (EFL) to further facilitate the uncertainty-guided self-training. Equipped with EFL, PT outperforms all previous baselines by a large margin and achieve new state-of-the-arts.