首页>
外国专利>
Voice recognition method and voice recognition device using voiced, unvoiced and silent section information
Voice recognition method and voice recognition device using voiced, unvoiced and silent section information
展开▼
机译:使用有声,无声和无声区间信息的语音识别方法和语音识别装置
展开▼
页面导航
摘要
著录项
相似文献
摘要
The present invention relates to a speech recognition method and a speech recognition device using voiced sound, unvoiced sound, and silence section information in a computer speech recognition application field, and a conventional speech recognition device extracts a feature vector for a speech signal in units of frames. After modeling by phoneme or similar phoneme unit, the ambiguity due to the increase of search time and the increase of recognition target language is increased by using the method of comparing the input unknown feature vector with the phoneme or similar phoneme model. Accordingly, the present invention further provides another feature parameter that can classify a voice signal into voiced / unvoiced / silent as well as existing feature parameters generally used for voice recognition among input signals when a human voice is input. By extracting the voiced, unvoiced, and silent sections correctly, the voiced / unvoiced / silent information is additionally used in the searcher of the speech recognizer to reduce the search target, which is more remarkable than the conventional speech recognizer. This reduces the search time and improves the search performance.
展开▼