Coupled Training for Multi-Source Domain Adaptation

Ohad Amosy, Gal Chechik

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

3 Scopus citations

Abstract

Unsupervised domain adaptation is often addressed by learning a joint representation of labeled samples from a source domain and unlabeled samples from a target domain. Unfortunately, hard sharing of representation may hurt adaptation because of negative transfer, where features that are useful for source domains are learned even if they hurt inference on the target domain. Here, we propose an alternative, soft sharing scheme. We train separate but weakly-coupled models for the source and the target data, while encouraging their predictions to agree. Training the two coupled models jointly effectively exploits the distribution over unlabeled target data and achieves high accuracy on the target. Specifically, we show analytically and empirically that the decision boundaries of the target model converge to low-density "valleys"of the target distribution. We evaluate our approach on four multi-source domain adaptation (MSDA) benchmarks, digits, amazon text reviews, Office-Caltech and images (DomainNet). We find that it consistently outperforms current MSDA SoTA, sometimes by a very large margin.

Original languageEnglish
Title of host publicationProceedings - 2022 IEEE/CVF Winter Conference on Applications of Computer Vision, WACV 2022
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1071-1080
Number of pages10
ISBN (Electronic)9781665409155
DOIs
StatePublished - 2022
Event22nd IEEE/CVF Winter Conference on Applications of Computer Vision, WACV 2022 - Waikoloa, United States
Duration: 4 Jan 20228 Jan 2022

Publication series

NameProceedings - 2022 IEEE/CVF Winter Conference on Applications of Computer Vision, WACV 2022

Conference

Conference22nd IEEE/CVF Winter Conference on Applications of Computer Vision, WACV 2022
Country/TerritoryUnited States
CityWaikoloa
Period4/01/228/01/22

Bibliographical note

Publisher Copyright:
© 2022 IEEE.

Keywords

  • Few-shot
  • Semi- and Un- supervised Learning
  • Transfer

Fingerprint

Dive into the research topics of 'Coupled Training for Multi-Source Domain Adaptation'. Together they form a unique fingerprint.

Cite this