A Style-Based Generator Architecture for Generative Adversarial Networks
We propose an alternative generator architecture for generative adversarial networks, borrowing from style transfer literature. The new architecture leads to an automatically learned, unsupervised separation of high-level attributes (e.g., pose and identity when trained on human faces) and stochasti...
Gespeichert in:
| Veröffentlicht in: | IEEE transactions on pattern analysis and machine intelligence 2021-12, Vol.43 (12), p.4217-4228 |
|---|---|
| Hauptverfasser: | , , |
| Format: | Artikel |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext bestellen |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | 4228 |
|---|---|
| container_issue | 12 |
| container_start_page | 4217 |
| container_title | IEEE transactions on pattern analysis and machine intelligence |
| container_volume | 43 |
| creator | Karras, Tero Laine, Samuli Aila, Timo |
| description | We propose an alternative generator architecture for generative adversarial networks, borrowing from style transfer literature. The new architecture leads to an automatically learned, unsupervised separation of high-level attributes (e.g., pose and identity when trained on human faces) and stochastic variation in the generated images (e.g., freckles, hair), and it enables intuitive, scale-specific control of the synthesis. The new generator improves the state-of-the-art in terms of traditional distribution quality metrics, leads to demonstrably better interpolation properties, and also better disentangles the latent factors of variation. To quantify interpolation quality and disentanglement, we propose two new, automated methods that are applicable to any generator architecture. Finally, we introduce a new, highly varied and high-quality dataset of human faces. |
| doi_str_mv | 10.1109/TPAMI.2020.2970919 |
| format | Article |
| fullrecord | <record><control><sourceid>proquest_RIE</sourceid><recordid>TN_cdi_proquest_journals_2592630570</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>8977347</ieee_id><sourcerecordid>2592630570</sourcerecordid><originalsourceid>FETCH-LOGICAL-c417t-b05ab333261449d125b90232eca4cb44fd4a83a1b554cbf47609302c5fcc88f03</originalsourceid><addsrcrecordid>eNpdkE1PwzAMhiMEgvHxB0BClbhw6XDipG2OZQKGxJcEnKM0dUWhWyFpQfv3dGxw4GTZfvzKehg75DDmHPTZ00N-ez0WIGAsdAqa6w024hp1jAr1JhsBT0ScZSLbYbshvAJwqQC32Q4K4AIARmyaR4_doqH43AYqoyuak7dd66Pcu5e6I9f1nqJqGKxX9SdFeflJPlhf2ya6o-6r9W9hn21Vtgl0sK577Pny4mkyjW_ur64n-U3sJE-7uABlC0QUCZdSl1yoQoNAQc5KV0hZldJmaHmh1NBXMk1AIwinKueyrALcY6er3HfffvQUOjOrg6OmsXNq-2AEKtAAmPABPfmHvra9nw_fGaG0SBBUugwUK8r5NgRPlXn39cz6heFglp7Nj2ez9GzWnoej43V0X8yo_Dv5FTsARyugJqK_dabTFGWK3zJZf8g</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2592630570</pqid></control><display><type>article</type><title>A Style-Based Generator Architecture for Generative Adversarial Networks</title><source>IEEE Xplore</source><creator>Karras, Tero ; Laine, Samuli ; Aila, Timo</creator><creatorcontrib>Karras, Tero ; Laine, Samuli ; Aila, Timo</creatorcontrib><description>We propose an alternative generator architecture for generative adversarial networks, borrowing from style transfer literature. The new architecture leads to an automatically learned, unsupervised separation of high-level attributes (e.g., pose and identity when trained on human faces) and stochastic variation in the generated images (e.g., freckles, hair), and it enables intuitive, scale-specific control of the synthesis. The new generator improves the state-of-the-art in terms of traditional distribution quality metrics, leads to demonstrably better interpolation properties, and also better disentangles the latent factors of variation. To quantify interpolation quality and disentanglement, we propose two new, automated methods that are applicable to any generator architecture. Finally, we introduce a new, highly varied and high-quality dataset of human faces.</description><identifier>ISSN: 0162-8828</identifier><identifier>EISSN: 1939-3539</identifier><identifier>EISSN: 2160-9292</identifier><identifier>DOI: 10.1109/TPAMI.2020.2970919</identifier><identifier>PMID: 32012000</identifier><identifier>CODEN: ITPIDJ</identifier><language>eng</language><publisher>United States: IEEE</publisher><subject>Aerospace electronics ; Convolution ; deep learning ; Generative adversarial networks ; Generative models ; Generators ; Image resolution ; Interpolation ; neural networks ; Training</subject><ispartof>IEEE transactions on pattern analysis and machine intelligence, 2021-12, Vol.43 (12), p.4217-4228</ispartof><rights>Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2021</rights><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c417t-b05ab333261449d125b90232eca4cb44fd4a83a1b554cbf47609302c5fcc88f03</citedby><cites>FETCH-LOGICAL-c417t-b05ab333261449d125b90232eca4cb44fd4a83a1b554cbf47609302c5fcc88f03</cites><orcidid>0000-0002-9437-4438 ; 0000-0002-0903-3197</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/8977347$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>314,777,781,793,27905,27906,54739</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/8977347$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc><backlink>$$Uhttps://www.ncbi.nlm.nih.gov/pubmed/32012000$$D View this record in MEDLINE/PubMed$$Hfree_for_read</backlink></links><search><creatorcontrib>Karras, Tero</creatorcontrib><creatorcontrib>Laine, Samuli</creatorcontrib><creatorcontrib>Aila, Timo</creatorcontrib><title>A Style-Based Generator Architecture for Generative Adversarial Networks</title><title>IEEE transactions on pattern analysis and machine intelligence</title><addtitle>TPAMI</addtitle><addtitle>IEEE Trans Pattern Anal Mach Intell</addtitle><description>We propose an alternative generator architecture for generative adversarial networks, borrowing from style transfer literature. The new architecture leads to an automatically learned, unsupervised separation of high-level attributes (e.g., pose and identity when trained on human faces) and stochastic variation in the generated images (e.g., freckles, hair), and it enables intuitive, scale-specific control of the synthesis. The new generator improves the state-of-the-art in terms of traditional distribution quality metrics, leads to demonstrably better interpolation properties, and also better disentangles the latent factors of variation. To quantify interpolation quality and disentanglement, we propose two new, automated methods that are applicable to any generator architecture. Finally, we introduce a new, highly varied and high-quality dataset of human faces.</description><subject>Aerospace electronics</subject><subject>Convolution</subject><subject>deep learning</subject><subject>Generative adversarial networks</subject><subject>Generative models</subject><subject>Generators</subject><subject>Image resolution</subject><subject>Interpolation</subject><subject>neural networks</subject><subject>Training</subject><issn>0162-8828</issn><issn>1939-3539</issn><issn>2160-9292</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2021</creationdate><recordtype>article</recordtype><sourceid>RIE</sourceid><recordid>eNpdkE1PwzAMhiMEgvHxB0BClbhw6XDipG2OZQKGxJcEnKM0dUWhWyFpQfv3dGxw4GTZfvzKehg75DDmHPTZ00N-ez0WIGAsdAqa6w024hp1jAr1JhsBT0ScZSLbYbshvAJwqQC32Q4K4AIARmyaR4_doqH43AYqoyuak7dd66Pcu5e6I9f1nqJqGKxX9SdFeflJPlhf2ya6o-6r9W9hn21Vtgl0sK577Pny4mkyjW_ur64n-U3sJE-7uABlC0QUCZdSl1yoQoNAQc5KV0hZldJmaHmh1NBXMk1AIwinKueyrALcY6er3HfffvQUOjOrg6OmsXNq-2AEKtAAmPABPfmHvra9nw_fGaG0SBBUugwUK8r5NgRPlXn39cz6heFglp7Nj2ez9GzWnoej43V0X8yo_Dv5FTsARyugJqK_dabTFGWK3zJZf8g</recordid><startdate>20211201</startdate><enddate>20211201</enddate><creator>Karras, Tero</creator><creator>Laine, Samuli</creator><creator>Aila, Timo</creator><general>IEEE</general><general>The Institute of Electrical and Electronics Engineers, Inc. (IEEE)</general><scope>97E</scope><scope>RIA</scope><scope>RIE</scope><scope>NPM</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7SC</scope><scope>7SP</scope><scope>8FD</scope><scope>JQ2</scope><scope>L7M</scope><scope>L~C</scope><scope>L~D</scope><scope>7X8</scope><orcidid>https://orcid.org/0000-0002-9437-4438</orcidid><orcidid>https://orcid.org/0000-0002-0903-3197</orcidid></search><sort><creationdate>20211201</creationdate><title>A Style-Based Generator Architecture for Generative Adversarial Networks</title><author>Karras, Tero ; Laine, Samuli ; Aila, Timo</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c417t-b05ab333261449d125b90232eca4cb44fd4a83a1b554cbf47609302c5fcc88f03</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2021</creationdate><topic>Aerospace electronics</topic><topic>Convolution</topic><topic>deep learning</topic><topic>Generative adversarial networks</topic><topic>Generative models</topic><topic>Generators</topic><topic>Image resolution</topic><topic>Interpolation</topic><topic>neural networks</topic><topic>Training</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Karras, Tero</creatorcontrib><creatorcontrib>Laine, Samuli</creatorcontrib><creatorcontrib>Aila, Timo</creatorcontrib><collection>IEEE All-Society Periodicals Package (ASPP) 2005-present</collection><collection>IEEE All-Society Periodicals Package (ASPP) 1998-Present</collection><collection>IEEE Xplore</collection><collection>PubMed</collection><collection>CrossRef</collection><collection>Computer and Information Systems Abstracts</collection><collection>Electronics & Communications Abstracts</collection><collection>Technology Research Database</collection><collection>ProQuest Computer Science Collection</collection><collection>Advanced Technologies Database with Aerospace</collection><collection>Computer and Information Systems Abstracts Academic</collection><collection>Computer and Information Systems Abstracts Professional</collection><collection>MEDLINE - Academic</collection><jtitle>IEEE transactions on pattern analysis and machine intelligence</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Karras, Tero</au><au>Laine, Samuli</au><au>Aila, Timo</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>A Style-Based Generator Architecture for Generative Adversarial Networks</atitle><jtitle>IEEE transactions on pattern analysis and machine intelligence</jtitle><stitle>TPAMI</stitle><addtitle>IEEE Trans Pattern Anal Mach Intell</addtitle><date>2021-12-01</date><risdate>2021</risdate><volume>43</volume><issue>12</issue><spage>4217</spage><epage>4228</epage><pages>4217-4228</pages><issn>0162-8828</issn><eissn>1939-3539</eissn><eissn>2160-9292</eissn><coden>ITPIDJ</coden><abstract>We propose an alternative generator architecture for generative adversarial networks, borrowing from style transfer literature. The new architecture leads to an automatically learned, unsupervised separation of high-level attributes (e.g., pose and identity when trained on human faces) and stochastic variation in the generated images (e.g., freckles, hair), and it enables intuitive, scale-specific control of the synthesis. The new generator improves the state-of-the-art in terms of traditional distribution quality metrics, leads to demonstrably better interpolation properties, and also better disentangles the latent factors of variation. To quantify interpolation quality and disentanglement, we propose two new, automated methods that are applicable to any generator architecture. Finally, we introduce a new, highly varied and high-quality dataset of human faces.</abstract><cop>United States</cop><pub>IEEE</pub><pmid>32012000</pmid><doi>10.1109/TPAMI.2020.2970919</doi><tpages>12</tpages><orcidid>https://orcid.org/0000-0002-9437-4438</orcidid><orcidid>https://orcid.org/0000-0002-0903-3197</orcidid></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | ISSN: 0162-8828 |
| ispartof | IEEE transactions on pattern analysis and machine intelligence, 2021-12, Vol.43 (12), p.4217-4228 |
| issn | 0162-8828 1939-3539 2160-9292 |
| language | eng |
| recordid | cdi_proquest_journals_2592630570 |
| source | IEEE Xplore |
| subjects | Aerospace electronics Convolution deep learning Generative adversarial networks Generative models Generators Image resolution Interpolation neural networks Training |
| title | A Style-Based Generator Architecture for Generative Adversarial Networks |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-17T15%3A20%3A08IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_RIE&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=A%20Style-Based%20Generator%20Architecture%20for%20Generative%20Adversarial%20Networks&rft.jtitle=IEEE%20transactions%20on%20pattern%20analysis%20and%20machine%20intelligence&rft.au=Karras,%20Tero&rft.date=2021-12-01&rft.volume=43&rft.issue=12&rft.spage=4217&rft.epage=4228&rft.pages=4217-4228&rft.issn=0162-8828&rft.eissn=1939-3539&rft.coden=ITPIDJ&rft_id=info:doi/10.1109/TPAMI.2020.2970919&rft_dat=%3Cproquest_RIE%3E2592630570%3C/proquest_RIE%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2592630570&rft_id=info:pmid/32012000&rft_ieee_id=8977347&rfr_iscdi=true |