. . "Navrhovan\u00FD projekt se zam\u011B\u0159uje na problematiku vizu\u00E1ln\u00ED synt\u00E9zy \u0159e\u010Di. Konkr\u00E9tn\u011B se \u0159e\u0161en\u00ED projektu bude soust\u0159edit na dopln\u011Bn\u00ED \u0159e\u010Dov\u00E9ho (akustick\u00E9ho) syntetiz\u00E9ru \u010De\u0161tiny o vizu\u00E1ln\u00ED synt\u00E9zu ve form\u011B animovan\u00E9ho syntetick\u00E9ho obli\u010Deje. C\u00EDlem je zlep\u0161en\u00EDporozum\u011Bn\u00ED \u010Dlov\u011Bkem v\u00A0hlu\u010Dn\u00E9m prost\u0159ed\u00ED nebo u lid\u00ED se sluchov\u00FDm posti\u017Een\u00EDm. Vlastn\u00ED vizu\u00E1ln\u00ED synt\u00E9za bude \u0159e\u0161ena metodou parametrick\u00E9ho modelu. \u0158e\u0161en\u00ED tohoto projektu zahrnuje celou \u0159adu d\u00EDl\u010D\u00EDch \u00FAloh. P\u0159edev\u0161\u00EDm je nutn\u00E9 shrom\u00E1\u017Edit nov\u00FD \u0159e\u010Dov\u00FD korpus,kter\u00FD bude obsahovat vizu\u00E1ln\u00ED z\u00E1znam podoby obli\u010Deje \u0159e\u010Dn\u00EDka b\u011Bhem namlouv\u00E1n\u00ED zvolen\u00E9ho textu. Mezi nejd\u016Fle\u017Eit\u011Bj\u0161\u00ED \u00FAlohy pat\u0159\u00ED volba parametrick\u00E9ho popisu tvaru obli\u010Deje a vytvo\u0159en\u00ED jeho parametrick\u00E9ho t\u0159\u00EDrozm\u011Brn\u00E9ho modelu, jeho\u017E tvar lze ovl\u00E1dat pomoc\u00EDzvolen\u00FDch parametr\u016F. Dal\u0161\u00ED \u00FAlohou je automatick\u00E1 segmentace obrazu a detekce v\u00FDznamn\u00FDch bod\u016F pro v\u00FDpo\u010Det hodnot zvolen\u00FDch parametr\u016F. S\u00A0t\u00EDm souvis\u00ED zp\u016Fsob vyb\u00EDr\u00E1n\u00ED reprezentativn\u00EDch \u0159e\u010Dov\u00FDch jednotek a vytvo\u0159en\u00ED vizu\u00E1ln\u00ED \u0159e\u010Dov\u00E9 datab\u00E1ze. Dal\u0161\u00ED navazuj\u00EDc\u00ED" . . "1"^^ . . . . . "During three-year period of the grant we solved the task of supplementing speech synthesizer with parametrically animable 3D model of a head that implements visual speech synthesis, that means imitate articulation of a real person. In the framework of th"@en . "0"^^ . "Proposed project focuses on the visual speech synthesis topic. The main aim of the project is to supplement present (auditory) speech synthesis system with the module of visual synthesis in the form of an animated synthetic face. The target is toincrease understanding by people in noisy environment or in case of hearing impairment. Solution of the task of visual synthesis will be based on the parametric model method. It involves several partial tasks. In the first place it is needed to collect anew speech corpus incorporating visual recording of a speaker face appearance during recording of selected text. The most important tasks are the choice of a face shape description and the design of a parametric three-dimensional model, which shape canbe controlled by the selected parameters. Next task is automatic image segmentation and important points detection for the computation of values of selected parameters. Related task is to find a way of selecting representative speech units from input"@en . . "0"^^ . . "B\u011Bhem t\u0159\u00EDlet\u00E9ho obdob\u00ED grantu jsme \u0159e\u0161ili \u00FAlohu dopln\u011Bn\u00ED \u0159e\u010Dov\u00E9ho syntetiz\u00E9ru o parametricky ovl\u00E1dan\u00FD animovateln\u00FD trojrozm\u011Brn\u00FD model hlavy, kter\u00FD realizuje takzvanou vizu\u00E1ln\u00ED synt\u00E9zu \u0159e\u010Di, tedy napodobuje pohyby \u00FAst a ostatn\u00EDch artikula\u010Dn\u00EDch org\u00E1n\u016F p\u0159i"@cs . . . "Neuvedeno."@en . . "13"^^ . "Vizu\u00E1ln\u00ED synt\u00E9za \u010De\u0161tiny metodou parametrick\u00E9ho modelu jako dopln\u011Bk \u0159e\u010Dov\u00E9ho syntetiz\u00E9ru" . "2009-01-15+01:00"^^ . "13"^^ . . . . . . "Visual Czech speech synthesis using parametric model method as a supplement of a speech synthesizer"@en . . . "GA102/03/0650" . "http://www.isvav.cz/projectDetail.do?rowId=GA102/03/0650"^^ .