. "Digital images form a large part of the Web content and proprietary image collections and their amount grows at enormous speed. Current techniques for searching in images are not satisfactory because search engines restrict image search to associated text or attribute annotations. Although content-based image retrieval techniques are improving with respect to effectiveness, they are still limited in the volume of images they can handle \u2013 the scalability is still a serious problem.This project aims at developing an innovative network technology platform which will allow organizing billions of images and searching by visual content. To enhance perceived quality of search results, specific techniques for clustering and ranking of results will also be developed. The platform will be tested on large image collections to demonstrate its properties. Several institutions and companies offering services such as online content-based retrieval, detection of plagiarism, filtering pornographic material, and brand reputation analysis based on logo detection, will cooperate as data providers."@en . "0"^^ . . "similarity" . "0"^^ . "1"^^ . . " content-based" . . . "GAP103/10/0886" . . "Content-based Image Retrieval on the Web Scale"@en . "Digit\u00E1ln\u00ED obr\u00E1zky dnes zab\u00EDraj\u00ED velikou \u010D\u00E1st webu, obzvl\u00E1\u0161t\u011B pak velmi rychle nar\u016Fst\u00E1 po\u010Det r\u016Fznorod\u00FDch kolekc\u00ED fotografi\u00ED. Ov\u0161em sou\u010Dasn\u00E9 techniky pro vyhled\u00E1v\u00E1n\u00ED obr\u00E1zk\u016F nejsou vyhovuj\u00EDc\u00ED, jeliko\u017E vyhled\u00E1va\u010De pracuj\u00ED pouze s textovou informaci spojenou s t\u011Bmito obr\u00E1zky. A\u010Dkoliv se neust\u00E1le objevuj\u00ED techniky zalo\u017Een\u00E9 na vyhled\u00E1v\u00E1n\u00ED obr\u00E1zk\u016F podle jejich vizu\u00E1ln\u00ED podoby, tak prozat\u00EDm nejsou pou\u017Eiteln\u00E9 pro velk\u00E9 mno\u017Estv\u00ED obr\u00E1zk\u016F \u2013 \u0161k\u00E1lovatelnost je st\u00E1le v\u00E1\u017En\u00FDm probl\u00E9mem.Tento projekt m\u00E1 za c\u00EDl vyvinout novou technologii, kter\u00E1 by umo\u017Enila organizovat miliardy obr\u00E1zk\u016F a z\u00E1rove\u0148 v nich vyhled\u00E1vat na z\u00E1klad\u011B jejich vizu\u00E1ln\u00ED podoby. Pro zlep\u0161en\u00ED v\u00FDsledn\u00E9 kvality nalezen\u00FDch obr\u00E1zk\u016F budou vyvinuty speci\u00E1ln\u00ED techniky pro jejich seskupov\u00E1n\u00ED a hodnocen\u00ED. Ve\u0161ker\u00E9 vlastnosti t\u00E9to technologie budou experiment\u00E1ln\u011B ov\u011B\u0159eny na rozs\u00E1hl\u00FDch datov\u00FDch mno\u017Ein\u00E1ch. Tyto obr\u00E1zky budou dod\u00E1ny od n\u011Bkolika instituc\u00ED zab\u00FDvaj\u00EDc\u00EDch se vizu\u00E1ln\u00EDm vyhled\u00E1v\u00E1n\u00EDm, odhalov\u00E1n\u00EDm plagi\u00E1t\u016F, filtrov\u00E1n\u00EDm pornografick\u00FDch materi\u00E1l\u016F nebo anal\u00FDzou renom\u00E9 firem zalo\u017Een\u00E9 na detekci obchodn\u00EDch zna\u010Dek." . "2014-01-30+01:00"^^ . "2010-01-01+01:00"^^ . "19"^^ . . "19"^^ . . . . . "Vizu\u00E1ln\u00ED vyhled\u00E1v\u00E1n\u00ED obr\u00E1zk\u016F na Webu" . " search" . " retrieval" . "2012-12-31+01:00"^^ . . " indexing" . "\u0158e\u0161itelsk\u00FD t\u00FDm vyvinul novou technologii pro obsahov\u00E9 vyhled\u00E1v\u00E1n\u00ED v obr\u00E1zc\u00EDch, kter\u00E1 \u00FAsp\u011B\u0161n\u011B kombinuje v\u00EDce ortogon\u00E1ln\u00EDch modalit. V\u00FDsledky jsou pops\u00E1ny ve 22 publikac\u00EDch, z nich\u017E 2 vy\u0161ly v \u010Dasopisech s IF nad 1,1."@cs . "http://www.isvav.cz/projectDetail.do?rowId=GAP103/10/0886"^^ . . . . "2012-03-30+02:00"^^ . . " distributed" . . "similarity; search; content-based; retrieval; distributed; indexing; scalability"@en . . "The project team developed advanced techniques for content-based image search that successfully combines several orotgonal modalities. The results have been described in 22 papers, 2 of them appeared in journals with IF over 1.1."@en .