Dual-Domain Teacher for Unsupervised Domain Adaptation Detection

doi:10.1109/TMM.2025.3535362

	Dual-Domain Teacher for Unsupervised Domain Adaptation Detection
	Wang, Fei 1; Zhao, Luhui 1; Hong, Shijie 1; Wang, Zhe 2; Liu, Chen 1; Gao, Changxin 3; Li, Jinsheng 4; Li, Xin 4; Luo, Dapeng 1
	2025
发表期刊	IEEE TRANSACTIONS ON MULTIMEDIA
ISSN	1520-9210
卷号	27 页码:4217-4226
摘要	Unsupervised domain adaptation for object detection aims to bridge the domain gap by transferring knowledge from a labeled source domain to an unlabeled target domain, thus improving the performance of detection models. Common strategies focus on aligning the feature distributions between source and target domains to reduce their discrepancies. However, achieving complete alignment is often not feasible in real-world situations due to a lack of annotations in the target domain. Recently, TeacherStudent approaches achieve feature alignment by generating reliable target pseudo-labels and become the dominant solution for addressing this issue. However, due to the domain shift, the teacher model bias to source domain, making it challenging to enhance the quality of target pseudo-labels. Some methods within this framework attempt to overcome the domain shift by incorporating distribution alignment components, yet these approaches also face challenges in achieving perfect alignment between domains. In this paper, we propose the Dual-Domain Teacher (DDT) method to address the domain adaptation detection problem by simultaneously detecting objects in both domains, thereby decreasing the need for perfect alignment. To address the issue of duplicate detection results produced by the Dual-Domain detection process, a candidate set refinement strategy is proposed to eliminate these duplicates across domains. Moreover, when teachers generate pseudo-labels by selecting reliable predictions with fixed confidence thresholds, valuable predictions may be overlooked in mutual learning. In our approach, a minimum variance-based dynamic threshold module is designed to mine valuable pseudo-labels by adaptively adjusting to the optimal threshold. Extensive experiments show that the DDT achieve a 56.7$\%$ mAP on the CityScapes-to-Foggy CityScapes task, marking a 4.8 point improvement over the latest methods. On the PASCAL VOC-to-Clipart1k task, our method reaches 51.2$\%$ mAP, outperforming previous state-of-the-art.
关键词	Adaptation models Training Reliability Object detection Detectors Annotations Translation Predictive models Feature extraction Representation learning Domain adaptation dual-domain teacher object detection pseudo-labels
DOI	10.1109/TMM.2025.3535362
收录类别	SCI
语种	英语
资助项目	National Natural Science Foundation of China[61302137] ; National Natural Science Foundation of China[62176097]
WOS研究方向	Computer Science ; Telecommunications
WOS类目	Computer Science, Information Systems ; Computer Science, Software Engineering ; Telecommunications
WOS记录号	WOS:001525570100011
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/41779
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Luo, Dapeng
作者单位	1.China Univ Geosci, Sch Mech Engn & Elect Informat, Wuhan 430074, Peoples R China 2.Chinese Acad Sci, Inst Comp Technol, Beijing 100045, Peoples R China 3.Huazhong Univ Sci & Technol, Sch Artificial Intelligence & Automat, Wuhan 430074, Peoples R China 4.Chinese Construct Third Engn Bur, Intelligent Technol Co Ltd, Wuhan 430070, Peoples R China
推荐引用方式 GB/T 7714	Wang, Fei,Zhao, Luhui,Hong, Shijie,et al. Dual-Domain Teacher for Unsupervised Domain Adaptation Detection[J]. IEEE TRANSACTIONS ON MULTIMEDIA,2025,27:4217-4226.
APA	Wang, Fei.,Zhao, Luhui.,Hong, Shijie.,Wang, Zhe.,Liu, Chen.,...&Luo, Dapeng.(2025).Dual-Domain Teacher for Unsupervised Domain Adaptation Detection.IEEE TRANSACTIONS ON MULTIMEDIA,27,4217-4226.
MLA	Wang, Fei,et al."Dual-Domain Teacher for Unsupervised Domain Adaptation Detection".IEEE TRANSACTIONS ON MULTIMEDIA 27(2025):4217-4226.