Rethinking the Truly Unsupervised Image-to-Image Translation

Cited 41 time in webofscience Cited 0 time in scopus
  • Hit : 90
  • Download : 0
DC FieldValueLanguage
dc.contributor.authorBaek, Kyungjuneko
dc.contributor.authorChoi, Yunjeyko
dc.contributor.authorUh, Youngjungko
dc.contributor.authorYoo, Jaejunko
dc.contributor.authorShim, Hyunjungko
dc.date.accessioned2022-08-23T09:00:15Z-
dc.date.available2022-08-23T09:00:15Z-
dc.date.created2022-07-07-
dc.date.created2022-07-07-
dc.date.issued2021-10-12-
dc.identifier.citation18th IEEE/CVF International Conference on Computer Vision, ICCV 2021, pp.14134 - 14143-
dc.identifier.urihttp://hdl.handle.net/10203/298044-
dc.description.abstractEvery recent image-to-image translation model inherently requires either image-level (i.e. input-output pairs) or set-level (i.e. domain labels) supervision. However, even set-level supervision can be a severe bottleneck for data collection in practice. In this paper, we tackle image-to-image translation in a fully unsupervised setting, i.e., neither paired images nor domain labels. To this end, we propose a truly unsupervised image-to-image translation model (TUNIT) that simultaneously learns to separate image domains and translates input images into the estimated domains. Experimental results show that our model achieves comparable or even better performance than the set-level supervised model trained with full labels, generalizes well on various datasets, and is robust against the choice of hyperparameters (e.g. the preset number of pseudo domains). Furthermore, TUNIT can be easily extended to semi-supervised learning with a few labeled data.-
dc.languageEnglish-
dc.publisherInstitute of Electrical and Electronics Engineers Inc.-
dc.titleRethinking the Truly Unsupervised Image-to-Image Translation-
dc.typeConference-
dc.identifier.wosid000798743204033-
dc.identifier.scopusid2-s2.0-85126500184-
dc.type.rimsCONF-
dc.citation.beginningpage14134-
dc.citation.endingpage14143-
dc.citation.publicationname18th IEEE/CVF International Conference on Computer Vision, ICCV 2021-
dc.identifier.conferencecountryCN-
dc.identifier.conferencelocationVirtual-
dc.identifier.doi10.1109/ICCV48922.2021.01389-
dc.contributor.localauthorShim, Hyunjung-
dc.contributor.nonIdAuthorBaek, Kyungjune-
dc.contributor.nonIdAuthorChoi, Yunjey-
dc.contributor.nonIdAuthorUh, Youngjung-
dc.contributor.nonIdAuthorYoo, Jaejun-
Appears in Collection
AI-Conference Papers(학술대회논문)
Files in This Item
There are no files associated with this item.
This item is cited by other documents in WoS
⊙ Detail Information in WoSⓡ Click to see webofscience_button
⊙ Cited 41 items in WoS Click to see citing articles in records_button

qr_code

  • mendeley

    citeulike


rss_1.0 rss_2.0 atom_1.0