Wasserstein Uncertainty Estimation for Adversarial Domain Matching

Rui Wang, Ruiyi Zhang, Ricardo Henao

Research output: Contribution to journalArticlepeer-review


Domain adaptation aims at reducing the domain shift between a labeled source domain and an unlabeled target domain, so that the source model can be generalized to target domains without fine tuning. In this paper, we propose to evaluate the cross-domain transferability between source and target samples by domain prediction uncertainty, which is quantified via Wasserstein gradient flows. Further, we exploit it for reweighting the training samples to alleviate the issue of domain shift. The proposed mechanism provides a meaningful curriculum for cross-domain transfer and adaptively rules out samples that contain too much domain specific information during domain adaptation. Experiments on several benchmark datasets demonstrate that our reweighting mechanism can achieve improved results in both balanced and partial domain adaptation.
Original languageEnglish (US)
JournalFrontiers in Big Data
StatePublished - May 10 2022
Externally publishedYes


Dive into the research topics of 'Wasserstein Uncertainty Estimation for Adversarial Domain Matching'. Together they form a unique fingerprint.

Cite this