Cross-domain Object Detection through Coarse-to-Fine Feature Adaptation

TL;DR

This paper introduces a coarse-to-fine feature adaptation method for cross-domain object detection, effectively transferring domain knowledge and achieving state-of-the-art results in various scenarios.

Contribution

It proposes a novel two-stage feature alignment approach using attention and adversarial learning for improved cross-domain detection.

Findings

Achieves state-of-the-art performance in cross-domain detection tasks.

Effective transfer of domain knowledge in foreground regions.

Broad applicability across different detection scenarios.

Abstract

Recent years have witnessed great progress in deep learning based object detection. However, due to the domain shift problem, applying off-the-shelf detectors to an unseen domain leads to significant performance drop. To address such an issue, this paper proposes a novel coarse-to-fine feature adaptation approach to cross-domain object detection. At the coarse-grained stage, different from the rough image-level or instance-level feature alignment used in the literature, foreground regions are extracted by adopting the attention mechanism, and aligned according to their marginal distributions via multi-layer adversarial learning in the common feature space. At the fine-grained stage, we conduct conditional distribution alignment of foregrounds by minimizing the distance of global prototypes with the same category but from different domains. Thanks to this coarse-to-fine feature adaptation, domain knowledge in foreground regions can be effectively transferred. Extensive experiments are carried out in various cross-domain detection scenarios. The results are state-of-the-art, which demonstrate the broad applicability and effectiveness of the proposed approach.