Beyond Entropy: Region Confidence Proxy for Wild Test-Time Adaptation

Zixuan Hu, Yichun Hu, Xiaotong Li, Shixiang Tang, Lingyu Duan
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:24371-24390, 2025.

Abstract

Wild Test-Time Adaptation (WTTA) is proposed to adapt a source model to unseen domains under extreme data scarcity and multiple shifts. Previous approaches mainly focused on sample selection strategies, while overlooking the fundamental problem on underlying optimization. Initially, we critically analyze the widely-adopted entropy minimization framework in WTTA and uncover its significant limitations in noisy optimization dynamics that substantially hinder adaptation efficiency. Through our analysis, we identify region confidence as a superior alternative to traditional entropy, however, its direct optimization remains computationally prohibitive for real-time applications. In this paper, we introduce a novel region-integrated method ReCAP that bypasses the lengthy process. Specifically, we propose a probabilistic region modeling scheme that flexibly captures semantic changes in embedding space. Subsequently, we develop a finite-to-infinite asymptotic approximation that transforms the intractable region confidence into a tractable and upper-bounded proxy. These innovations significantly unlock the overlooked potential dynamics in local region in a concise solution. Our extensive experiments demonstrate the consistent superiority of ReCAP over existing methods across various datasets and wild scenarios. The source code will be available at https://github.com/hzcar/ReCAP.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-hu25i, title = {Beyond Entropy: Region Confidence Proxy for Wild Test-Time Adaptation}, author = {Hu, Zixuan and Hu, Yichun and Li, Xiaotong and Tang, Shixiang and Duan, Lingyu}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {24371--24390}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/hu25i/hu25i.pdf}, url = {https://proceedings.mlr.press/v267/hu25i.html}, abstract = {Wild Test-Time Adaptation (WTTA) is proposed to adapt a source model to unseen domains under extreme data scarcity and multiple shifts. Previous approaches mainly focused on sample selection strategies, while overlooking the fundamental problem on underlying optimization. Initially, we critically analyze the widely-adopted entropy minimization framework in WTTA and uncover its significant limitations in noisy optimization dynamics that substantially hinder adaptation efficiency. Through our analysis, we identify region confidence as a superior alternative to traditional entropy, however, its direct optimization remains computationally prohibitive for real-time applications. In this paper, we introduce a novel region-integrated method ReCAP that bypasses the lengthy process. Specifically, we propose a probabilistic region modeling scheme that flexibly captures semantic changes in embedding space. Subsequently, we develop a finite-to-infinite asymptotic approximation that transforms the intractable region confidence into a tractable and upper-bounded proxy. These innovations significantly unlock the overlooked potential dynamics in local region in a concise solution. Our extensive experiments demonstrate the consistent superiority of ReCAP over existing methods across various datasets and wild scenarios. The source code will be available at https://github.com/hzcar/ReCAP.} }
Endnote
%0 Conference Paper %T Beyond Entropy: Region Confidence Proxy for Wild Test-Time Adaptation %A Zixuan Hu %A Yichun Hu %A Xiaotong Li %A Shixiang Tang %A Lingyu Duan %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-hu25i %I PMLR %P 24371--24390 %U https://proceedings.mlr.press/v267/hu25i.html %V 267 %X Wild Test-Time Adaptation (WTTA) is proposed to adapt a source model to unseen domains under extreme data scarcity and multiple shifts. Previous approaches mainly focused on sample selection strategies, while overlooking the fundamental problem on underlying optimization. Initially, we critically analyze the widely-adopted entropy minimization framework in WTTA and uncover its significant limitations in noisy optimization dynamics that substantially hinder adaptation efficiency. Through our analysis, we identify region confidence as a superior alternative to traditional entropy, however, its direct optimization remains computationally prohibitive for real-time applications. In this paper, we introduce a novel region-integrated method ReCAP that bypasses the lengthy process. Specifically, we propose a probabilistic region modeling scheme that flexibly captures semantic changes in embedding space. Subsequently, we develop a finite-to-infinite asymptotic approximation that transforms the intractable region confidence into a tractable and upper-bounded proxy. These innovations significantly unlock the overlooked potential dynamics in local region in a concise solution. Our extensive experiments demonstrate the consistent superiority of ReCAP over existing methods across various datasets and wild scenarios. The source code will be available at https://github.com/hzcar/ReCAP.
APA
Hu, Z., Hu, Y., Li, X., Tang, S. & Duan, L.. (2025). Beyond Entropy: Region Confidence Proxy for Wild Test-Time Adaptation. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:24371-24390 Available from https://proceedings.mlr.press/v267/hu25i.html.

Related Material