ESBN: Estimation Shift of Batch Normalization for Source-free Universal Domain Adaptation

Li, Jiao; Su, Houcheng; Wang, Bingli; Min, Yuandong; Wang, Mengzhu; Yin, Nan; Wang, Shanshan; Guo, Jingcai

doi:10.24963/ijcai.2025/155

ESBN: Estimation Shift of Batch Normalization for Source-free Universal Domain Adaptation

Jiao Li, Houcheng Su, Bingli Wang, Yuandong Min, Mengzhu Wang, Nan Yin, Shanshan Wang, Jingcai Guo

Download BibTeX Cite

Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence. Main Track. Pages 1386-1394.

https://doi.org/10.24963/ijcai.2025/155

Keywords

Computer Vision: CV: Transfer, low-shot, semi- and un- supervised learning
Computer Vision: CV: Recognition (object detection, categorization)

Abstract

Domain adaptation (DA) is crucial for transferring models trained in one domain to perform well in a different, often unseen domain. Traditional methods, including unsupervised domain adaptation (UDA) and source-free domain adaptation (SFDA), have made significant progress. However, most existing DA methods rely heavily on Batch Normalization (BN) layers, which are not optimal in source-free settings, where the source domain is unavailable for comparison. In this study, we propose a novel method, ESBN, which addresses the challenge of domain shift by adjusting the placement of normalization layers and replacing BN with Batch-free Normalization (BFN). Unlike BN, BFN is less dependent on batch statistics and provides more robust feature representations through instance-specific statistics. We systematically investigate the effects of different BN layer placements across various network configurations and demonstrate that selective replacement with BFN improves generalization performance. Extensive experiments on multiple domain adaptation benchmarks show that our approach outperforms state-of-the-art methods, particularly in challenging scenarios such as Open-Partial Domain Adaptation (OPDA).

Jiao Li, Houcheng Su, Bingli Wang, Yuandong Min, Mengzhu Wang, Nan Yin, Shanshan Wang, and Jingcai Guo. 2025. ESBN: Estimation Shift of Batch Normalization for Source-free Universal Domain Adaptation. In Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence, IJCAI-25. IJCAI, 1386–1394. https://doi.org/10.24963/ijcai.2025/155

@inproceedings{ijcai2025p155,
  title     = {ESBN: Estimation Shift of Batch Normalization for Source-free Universal Domain Adaptation},
  author    = {Li, Jiao and Su, Houcheng and Wang, Bingli and Min, Yuandong and Wang, Mengzhu and Yin, Nan and Wang, Shanshan and Guo, Jingcai},
  booktitle = {Proceedings of the Thirty-Fourth International Joint Conference on
               Artificial Intelligence, {IJCAI-25}},
  publisher = {International Joint Conferences on Artificial Intelligence Organization},
  editor    = {James Kwok},
  pages     = {1386--1394},
  year      = {2025},
  month     = {8},
  note      = {Main Track},
  doi       = {10.24963/ijcai.2025/155},
  url       = {https://doi.org/10.24963/ijcai.2025/155},
}

TY  - CPAPER
TI  - ESBN: Estimation Shift of Batch Normalization for Source-free Universal Domain Adaptation
AU  - Li, Jiao
AU  - Su, Houcheng
AU  - Wang, Bingli
AU  - Min, Yuandong
AU  - Wang, Mengzhu
AU  - Yin, Nan
AU  - Wang, Shanshan
AU  - Guo, Jingcai
T2  - Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence, IJCAI-25
PB  - International Joint Conferences on Artificial Intelligence Organization
PY  - 2025
SP  - 1386
EP  - 1394
DO  - 10.24963/ijcai.2025/155
UR  - https://doi.org/10.24963/ijcai.2025/155
ER  -