"2014-04-07+02:00"^^ . . . "Centrum pro multi-mod\u00E1ln\u00ED interpretaci dat velk\u00E9ho rozsahu" . "Projekt m\u00E1 za c\u00EDl vyu\u017E\u00EDt velmi rozs\u00E1hl\u00E1 neanotovan\u00E1 multimod\u00E1ln\u00ED data, p\u0159ev\u00E1\u017En\u011B videoz\u00E1znamy, k v\u00FDznamn\u00E9mu posunu stavu pozn\u00E1n\u00ED v oblasti porozum\u011Bn\u00ED, interpretace, anotace a vyhled\u00E1v\u00E1n\u00ED obrazov\u00FDch sekvenc\u00ED, \u0159e\u010Dov\u00E9ho sign\u00E1lu a p\u0159irozen\u00E9ho jazyka a to kombinac\u00ED metod zalo\u017Een\u00FDch na u\u010Den\u00ED bez u\u010Ditele. Zam\u011B\u0159\u00ED se na probl\u00E9my, kter\u00E9 jsou z pohledu jedn\u00E9 modality velmi t\u011B\u017Eko \u0159e\u0161iteln\u00E9 a vyu\u017Eije interdisciplin\u00E1rn\u00EDho p\u0159\u00EDstupu. Pokrok v jednotliv\u00FDch oborech (vid\u011Bn\u00ED, \u0159e\u010D, jazyk) bude dosa\u017Een kombinac\u00ED metoda a zejm\u00E9na k\u0159\u00ED\u017Eov\u00FDm u\u010Den\u00EDm, kdy v\u00FDstup jedn\u00E9 z modalit bude zdrojem tr\u00E9novac\u00EDch dat druh\u00FDch modalit. Za \u00FA\u010Delem efektivn\u00EDho zpracov\u00E1n\u00ED velk\u00FDch soubor\u016F dat v\u0161ech modalit, se projekt tak\u00E9 zam\u011B\u0159\u00ED na obecn\u00E9 probl\u00E9my organizace, indexace a vyhled\u00E1v\u00E1n\u00ED na z\u00E1klad\u011B podobnosti, co\u017E m\u00E1 velk\u00FD v\u00FDznam p\u0159i budov\u00E1n\u00ED skute\u010Dn\u00FDch aplikac\u00ED. Konsorcium se skl\u00E1d\u00E1 z mezin\u00E1rodn\u011B uzn\u00E1van\u00FDch pracovi\u0161\u0165, dosahuj\u00EDc\u00EDch \u0161pi\u010Dkov\u00E9 v\u00FDsledky. V projektu. dojde k p\u0159enosu znalost\u00ED, metodologi\u00ED a \u0159e\u0161en\u00FDch probl\u00E9m\u016F. Krom\u011B v\u011Bdeck\u00FDch v\u00FDstup\u016F ve form\u011B publikac\u00ED budou v\u00FDstupem projektu dva demonstr\u00E1tory" . . "78"^^ . "multimodal data understanding video annotation information extraction information retrieval natu"@en . "78"^^ . . "2018-12-31+01:00"^^ . "The project aims at exploiting large collections of unlabeled multi-modal data, mainly video footage, to further state-of-the-art in video, audio and natural language understanding, interpretation, annotation and retrieval by combining unsupervised and semi-supervised learning. It will address problems that are very difficult (some probably impossible) to solve in a single modality by adopting an interdisciplinary approach. Progress in individual areas - vision, language and speech will be achieved by co-training and by exploiting results of other modalities as cross-training data. For efficient processing of large data collections, the project will also concern generic problems of organization, indexing, and searching based on similarity that is critical for building real-life applications. The consortium comprises internationally-recognized groups possessing cutting edge expertise in the research areas. The project will benefit from sharing of expertise, data, and methodologies. Besides scientific results presented in publications, two demonstrators will be produced."@en . . . "Center for Large Scale Multi-modal Data Interpretation"@en . "2012-01-01+01:00"^^ . "1"^^ . . "3"^^ . . . . . . . . . . "2015-02-09+01:00"^^ . "0"^^ . "GBP103/12/G084" . . . . "http://www.isvav.cz/projectDetail.do?rowId=GBP103/12/G084"^^ .