Prosody based endpoint detection

   
   

A method and apparatus are provided for performing prosody based endpoint detection of speech in a speech recognition system. Input speech represents an utterance, which has an intonation pattern. An end-of-utterance condition is identified based on prosodic parameters of the utterance, such as the intonation pattern and the duration of the final syllable of the utterance, as well as non-prosodic parameters, such as the log energy of the speech.

Метод и прибор обеспечены для выполнять prosody основанное обнаружение endpoint речи в системе опознавания речи. Речь входного сигнала представляет произнесение, которое имеет картину интонирования. Определено условие конц-$$$-PROIZNESENI4 основало на стиховедческих параметрах произнесения, such as картина интонирования и продолжительность окончательного силлабула произнесения, также,как нон-stixovedceskie параметры, such as энергия журнала речи.

 
Web www.patentalert.com

< Load-adjusted speech recognition

< Distributed voice web architecture and associated components and methods

> Recursively excited linear prediction speech coder

> Performing actions identified in recognized speech

~ 00172