COSDA: Counterfactual-based Susceptibility Risk Framework for Open-Set Domain Adaptation

Wenxu Wang, Rui Zhou, Jing Wang, Yun Zhou, Cheng Zhu, Ruichun Tang, Bo Han, Nevin L. Zhang
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:65845-65866, 2025.

Abstract

Open-Set Domain Adaptation (OSDA) aims to transfer knowledge from the labeled source domain to the unlabeled target domain that contains unknown categories, thus facing the challenges of domain shift and unknown category recognition. While recent works have demonstrated the potential of causality for domain alignment, little exploration has been conducted on causal-inspired theoretical frameworks for OSDA. To fill this gap, we introduce the concept of Susceptibility and propose a novel Counterfactual-based susceptibility risk framework for OSDA, termed COSDA. Specifically, COSDA consists of three novel components: (i) a Susceptibility Risk Estimator (SRE) for capturing causal information, along with comprehensive derivations of the computable theoretical upper bound, forming a risk minimization framework under the OSDA paradigm; (ii) a Contrastive Feature Alignment (CFA) module, which is theoretically proven based on mutual information to satisfy the Exogeneity assumption and facilitate cross-domain feature alignment; (iii) a Virtual Multi-unknown-categories Prototype (VMP) pseudo-labeling strategy, providing label information by measuring how similar samples are to known and multiple virtual unknown category prototypes, thereby assisting in open-set recognition and intra-class discriminative feature learning. Extensive experiments demonstrate that our approach achieves state-of-the-art performance.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-wang25fd, title = {{COSDA}: Counterfactual-based Susceptibility Risk Framework for Open-Set Domain Adaptation}, author = {Wang, Wenxu and Zhou, Rui and Wang, Jing and Zhou, Yun and Zhu, Cheng and Tang, Ruichun and Han, Bo and Zhang, Nevin L.}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {65845--65866}, 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/wang25fd/wang25fd.pdf}, url = {https://proceedings.mlr.press/v267/wang25fd.html}, abstract = {Open-Set Domain Adaptation (OSDA) aims to transfer knowledge from the labeled source domain to the unlabeled target domain that contains unknown categories, thus facing the challenges of domain shift and unknown category recognition. While recent works have demonstrated the potential of causality for domain alignment, little exploration has been conducted on causal-inspired theoretical frameworks for OSDA. To fill this gap, we introduce the concept of Susceptibility and propose a novel Counterfactual-based susceptibility risk framework for OSDA, termed COSDA. Specifically, COSDA consists of three novel components: (i) a Susceptibility Risk Estimator (SRE) for capturing causal information, along with comprehensive derivations of the computable theoretical upper bound, forming a risk minimization framework under the OSDA paradigm; (ii) a Contrastive Feature Alignment (CFA) module, which is theoretically proven based on mutual information to satisfy the Exogeneity assumption and facilitate cross-domain feature alignment; (iii) a Virtual Multi-unknown-categories Prototype (VMP) pseudo-labeling strategy, providing label information by measuring how similar samples are to known and multiple virtual unknown category prototypes, thereby assisting in open-set recognition and intra-class discriminative feature learning. Extensive experiments demonstrate that our approach achieves state-of-the-art performance.} }
Endnote
%0 Conference Paper %T COSDA: Counterfactual-based Susceptibility Risk Framework for Open-Set Domain Adaptation %A Wenxu Wang %A Rui Zhou %A Jing Wang %A Yun Zhou %A Cheng Zhu %A Ruichun Tang %A Bo Han %A Nevin L. Zhang %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-wang25fd %I PMLR %P 65845--65866 %U https://proceedings.mlr.press/v267/wang25fd.html %V 267 %X Open-Set Domain Adaptation (OSDA) aims to transfer knowledge from the labeled source domain to the unlabeled target domain that contains unknown categories, thus facing the challenges of domain shift and unknown category recognition. While recent works have demonstrated the potential of causality for domain alignment, little exploration has been conducted on causal-inspired theoretical frameworks for OSDA. To fill this gap, we introduce the concept of Susceptibility and propose a novel Counterfactual-based susceptibility risk framework for OSDA, termed COSDA. Specifically, COSDA consists of three novel components: (i) a Susceptibility Risk Estimator (SRE) for capturing causal information, along with comprehensive derivations of the computable theoretical upper bound, forming a risk minimization framework under the OSDA paradigm; (ii) a Contrastive Feature Alignment (CFA) module, which is theoretically proven based on mutual information to satisfy the Exogeneity assumption and facilitate cross-domain feature alignment; (iii) a Virtual Multi-unknown-categories Prototype (VMP) pseudo-labeling strategy, providing label information by measuring how similar samples are to known and multiple virtual unknown category prototypes, thereby assisting in open-set recognition and intra-class discriminative feature learning. Extensive experiments demonstrate that our approach achieves state-of-the-art performance.
APA
Wang, W., Zhou, R., Wang, J., Zhou, Y., Zhu, C., Tang, R., Han, B. & Zhang, N.L.. (2025). COSDA: Counterfactual-based Susceptibility Risk Framework for Open-Set Domain Adaptation. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:65845-65866 Available from https://proceedings.mlr.press/v267/wang25fd.html.

Related Material