Exploring Inter-Domain Wasserstein Metric for Adaptive Object Detection
Top Cited Papers
- 27 February 2026
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. PP (15209210) , 1-11
Abstract
Cross-domain adaptation has achieved significant development in recent years. Nevertheless, the models' performance varies dramatically across different scenarios. How to quantitatively measure inter-domain discrepancies to guide model training remains a challenging problem. Existing methods mainly focus on the trivial pixel-wise differences between the cross domain images, while they ignore the holistic discrepancies of the domain-specific distributions in various scenarios. Thus their effectivenesses are greatly limited in realistic applications. In this paper, we propose a universal method for measuring inter domain discrepancies based on Wasserstein distance. It alleviates the impact of intra-domain discrepancies on measurements and enables precise and quantitative representation of inter-domain discrepancies. We further integrate this metric into the generative models, and propose an assistant domain to conduct domain knowledge transfer for cross-domain object detection task. Experiments on three benchmarks validate the effectiveness of the proposed inter-domain measurement metric. Specifically, we achieve 51.8% mAP on CityScapes, 45.3% mAP on Clipart and 58.9% mAP on Watercolor, which are 1.5%, 0.5% and 0.8% higher than the state-of-the-art schemes, respectively.Keywords
Funding Information
- National Key Research and Development Program of China (2024YFC3308500)
- National Natural Science Foundation of China (62406036)
- State Key Laboratory of Networking and Switching Technology (NST20250110)
- SMP Zhipu.AI Large Model Cross Disciplinary Fund (ZPCG20241029322)
This publication has 37 references indexed in Scilit:
- MTTrans: Cross-domain Object Detection with Mean Teacher Transformer2022
- Contrastive Learning for Unpaired Image-to-Image Translation2020
- End-to-End Object Detection with Transformers2020
- AnimeGAN: A Novel Lightweight GAN for Photo Animation2020
- Semantic Foggy Scene Understanding with Synthetic DataInternational Journal of Computer Vision, 2018
- Optimal Transport for Diffeomorphic Registration2017
- The Cityscapes Dataset for Semantic Urban Scene Understanding2016
- Optimal Transport2009
- A new weighted metric: the relative metric IJournal of Mathematical Analysis and Applications, 2002
- A remark on M. M. Day’s characterization of inner-product spaces and a conjecture of L. M. BlumenthalProceedings of the American Mathematical Society, 1952