Artykuł w czasopiśmie
Brak miniatury
Licencja
Dostęp zamkniętyG3AN: Disentangling Appearance and Motion for Video Generation
| dc.abstract.en | Creating realistic human videos entails the challenge of being able to simultaneously generate both appearance, as well as motion. To tackle this challenge, we introduce G3AN, a novel spatio-temporal generative model, which seeks to capture the distribution of high dimensional video data and to model appearance and motion in disentangled manner. The latter is achieved by decomposing appearance and motion in a three-stream Generator, where the main stream aims to model spatio-temporal consistency, whereas the two auxiliary streams augment the main stream with multi-scale appearance and motion features, respectively. An extensive quantitative and qualitative analysis shows that our model systematically and significantly outperforms state-of-the-art methods on the facial expression datasets MUG and UvA-NEMO, as well as the Weizmann and UCF101 datasets on human action. Additional analysis on the learned latent representations confirms the successful decomposition of appearance and motion. |
| dc.affiliation | Uniwersytet Warszawski |
| dc.conference.country | Stany Zjednoczone |
| dc.conference.datefinish | 2020-06-19 |
| dc.conference.datestart | 2020-06-14 |
| dc.conference.place | Seattle |
| dc.conference.series | IEEE Conference on Computer Vision and Pattern Recognition |
| dc.conference.series | IEEE Conference on Computer Vision and Pattern Recognition |
| dc.conference.seriesshortcut | CVPR |
| dc.conference.shortcut | CVPR 2020 |
| dc.conference.weblink | http://cvpr2020.thecvf.com/ |
| dc.contributor.author | Wang, Yaohui |
| dc.contributor.author | Dantcheva, Antitza |
| dc.contributor.author | Bremond, Francois |
| dc.contributor.author | Biliński, Piotr |
| dc.date.accessioned | 2024-01-25T01:42:45Z |
| dc.date.available | 2024-01-25T01:42:45Z |
| dc.date.issued | 2020 |
| dc.description.finance | Publikacja bezkosztowa |
| dc.identifier.doi | 10.1109/CVPR42600.2020.00531 |
| dc.identifier.uri | https://repozytorium.uw.edu.pl//handle/item/107674 |
| dc.identifier.weblink | http://xplorestaging.ieee.org/ielx7/9142308/9156271/09157816.pdf?arnumber=9157816 |
| dc.language | eng |
| dc.pbn.affiliation | computer and information sciences |
| dc.relation.pages | 1-10 |
| dc.rights | ClosedAccess |
| dc.sciencecloud | nosend |
| dc.title | G3AN: Disentangling Appearance and Motion for Video Generation |
| dc.type | JournalArticle |
| dspace.entity.type | Publication |