"[3983A94FD0EC]" . "Modelling of non-speech events brings the necessary robustness in the recognition of natural or spontaneous utterances which are usually full of such acoustic disfluencies. This paper presents the solution of speaker non-speech event modelling commonly with the analyzes how efficiently these events are modelled. Firstly, the procedure for efficient training of non-speech event models on read speech data is presented. The results of experiments with simple ASR achieved 26\\,\\% decrease of word error rate and a~significant decrease of insertion rate with these models. Secondly, the extension of training data with spontaneous speech collection is described. It contributes to the availability of more natural data for training purposes and mainly to the better training of non-speech event models, which is demonstrated by the experiment on filled pause recognition."@en . . "P(GA102/08/0707), P(GD102/08/H008), Z(MSM6840770014)" . "\u017Dilina" . "Speaker Non-Speech Event Modelling in Recognition of Read and Spontaneous Speech"@en . "\u017Dilina" . . "978-80-8070-953-2" . "RIV/68407700:21230/08:03149357" . "Speaker Non-Speech Event Modelling in Recognition of Read and Spontaneous Speech" . . . . . . . "2008-11-20+01:00"^^ . . "2"^^ . . "Modelov\u00E1n\u00ED ne\u0159e\u010Dov\u00FDch ud\u00E1lost\u00ED p\u0159i rozpozn\u00E1v\u00E1n\u00ED \u010Dten\u00FDch a spont\u00E1nn\u00EDch promluv"@cs . . "RIV/68407700:21230/08:03149357!RIV09-MSM-21230___" . . "Modelov\u00E1n\u00ED ne\u0159e\u010Dov\u00FDch ud\u00E1lost\u00ED p\u0159i rozpozn\u00E1v\u00E1n\u00ED \u010Dten\u00FDch a spont\u00E1nn\u00EDch promluv"@cs . . . "Rajnoha, Josef" . "\u017Dilinsk\u00E1 univerzita v \u017Diline. Elektrotechnick\u00E1 fakulta" . . "6"^^ . "Modelov\u00E1n\u00ED ne\u0159e\u010Dov\u00FDch ud\u00E1lost\u00ED umo\u017E\u0148uje zv\u00FD\u0161it robustnost syst\u00E9mu v p\u0159\u00EDpad\u011B rozpozn\u00E1v\u00E1n\u00ED p\u0159irozen\u00E9 \u010Di spont\u00E1nn\u00ED promluvy, kter\u00E9 se vyzna\u010Duj\u00ED vy\u0161\u0161\u00EDm v\u00FDskytem t\u011Bchto ud\u00E1lost\u00ED. \u010Cl\u00E1nek popisuje \u0159e\u0161en\u00ED \u00FAlohy modelov\u00E1n\u00ED ne\u0159e\u010Dov\u00FDch ud\u00E1lost\u00ED \u0159e\u010Dn\u00EDka a p\u0159in\u00E1\u0161\u00ED anal\u00FDzu, jak efektivn\u011B lze tyto ud\u00E1losti modelovat. V prvn\u00ED \u010D\u00E1sti je prezentov\u00E1n postup tr\u00E9nov\u00E1n\u00ED model\u016F ne\u0159e\u010Dov\u00FDch ud\u00E1lost\u00ED na \u010Dten\u00FDch promluv\u00E1ch. Tento postup p\u0159in\u00E1\u0161\u00ED sn\u00ED\u017Een\u00ED chybovosti a\u017E o 26% a v\u00FDrazn\u00E9 sn\u00ED\u017Een\u00ED chybn\u011B vkl\u00E1dan\u00FDch slov. Druh\u00E1 \u010D\u00E1st se zab\u00FDv\u00E1 roz\u0161\u00ED\u0159en\u00EDm tr\u00E9novac\u00ED sady dat o spont\u00E1nn\u00ED promluvy, kter\u00E9 napom\u00E1h\u00E1 kvalitn\u011Bj\u0161\u00EDmu natr\u00E9nov\u00E1n\u00ED model\u016F, co\u017E je dolo\u017Eeno na experimentu s modelov\u00E1n\u00EDm v\u00E1hav\u00FDch pauz."@cs . "2"^^ . "Speaker Non-Speech Event Modelling in Recognition of Read and Spontaneous Speech"@en . . "Poll\u00E1k, Petr" . "non-speech event modelling; non-speech events; speech recognition; spontaneous speech"@en . . . . "396308" . "Digital Technologies 2008" . "Modelling of non-speech events brings the necessary robustness in the recognition of natural or spontaneous utterances which are usually full of such acoustic disfluencies. This paper presents the solution of speaker non-speech event modelling commonly with the analyzes how efficiently these events are modelled. Firstly, the procedure for efficient training of non-speech event models on read speech data is presented. The results of experiments with simple ASR achieved 26\\,\\% decrease of word error rate and a~significant decrease of insertion rate with these models. Secondly, the extension of training data with spontaneous speech collection is described. It contributes to the availability of more natural data for training purposes and mainly to the better training of non-speech event models, which is demonstrated by the experiment on filled pause recognition." . . "Speaker Non-Speech Event Modelling in Recognition of Read and Spontaneous Speech" . "21230" . .