Multi-domain Adversarial Variational Bayesian Inference for Domain Generalization

Simple item page
dc.contributor.authorGao, Zhifan
dc.contributor.authorGuo, Saidi
dc.contributor.authorXu, Chenchu
dc.contributor.authorZhang, Jinglin
dc.contributor.authorGong, Mingming
dc.contributor.authorDel Ser, Javier
dc.contributor.authorLi, Shuo
dc.contributor.institutionIA
dc.date.accessioned2024-07-24T12:11:19Z
dc.date.available2024-07-24T12:11:19Z
dc.date.issued2022
dc.descriptionPublisher Copyright: IEEE
dc.description.abstractDomain generalization aims to learn common knowledge from multiple observed source domains and transfer it to unseen target domains, e.g. the object recognition in varieties of visual environments. Traditional domain generalization methods aim to learn the feature representation of the raw data with its distribution invariant across domains. This relies on the assumption that the two posterior distributions (the distributions of the label given the feature distribution and given the raw data) are stable in different domains. However, this does not always hold in many practical situations. In this paper, we relax the above assumption by permitting the posterior distribution of the label given the raw data changes in difference domains, and thus focuses on a more realistic learning problem that infers the conditional domain-invariant feature representation. Specifically, a multi-domain adversarial variational Bayesian inference approach is proposed to minimize the inter-domain discrepancy of the conditional distributions of the feature given the label. Besides, it is imposed by the constraints from the adversarial learning and feedback mechanism to enhance the condition invariant feature representation. The extensive experiments on two datasets demonstrate the effectiveness of our approach, as well as the state-of-the-art performance comparing with thirteen methods.en
dc.description.statusPeer reviewed
dc.format.extent1
dc.identifier.citationGao , Z , Guo , S , Xu , C , Zhang , J , Gong , M , Del Ser , J & Li , S 2022 , ' Multi-domain Adversarial Variational Bayesian Inference for Domain Generalization ' , IEEE Transactions on Circuits and Systems for Video Technology , pp. 1 . https://doi.org/10.1109/TCSVT.2022.3232112
dc.identifier.doi10.1109/TCSVT.2022.3232112
dc.identifier.issn1051-8215
dc.identifier.urihttps://hdl.handle.net/11556/4171
dc.identifier.urlhttp://www.scopus.com/inward/record.url?scp=85146234522&partnerID=8YFLogxK
dc.language.isoeng
dc.relation.ispartofIEEE Transactions on Circuits and Systems for Video Technology
dc.rightsinfo:eu-repo/semantics/restrictedAccess
dc.subject.keywordsAdaptation models
dc.subject.keywordsAdversarial machine learning
dc.subject.keywordsBayes methods
dc.subject.keywordsBiomedical imaging
dc.subject.keywordsDomain generalization
dc.subject.keywordsTask analysis
dc.subject.keywordsTraining
dc.subject.keywordsVisualization
dc.subject.keywordsvariational auto-encoder
dc.subject.keywordsvisual object recognition
dc.subject.keywordsMedia Technology
dc.subject.keywordsElectrical and Electronic Engineering
dc.titleMulti-domain Adversarial Variational Bayesian Inference for Domain Generalizationen
dc.typejournal article
Files