Title :
Disquisition of Speech Recognition In VDR
Author :
Zhou, Wenjun ; Hao, Yanling
Author_Institution :
Coll. of Autom., Harbin Eng. Univ.
Abstract :
An speech recognition based on the energy difference degree (EDD) argument was proposed. Because voiced sound has distinct formants, while great noise with impact has not, therefore, this algorithm greatly improves the capturing capability of it. By setting up and down threshold of EDD parameter, weak and loud noise was disposed and results was obtained better. The analysis in theory and the result of emulation shows that this method has preferable effect under low Signal-to-Noise used in speech recognition in VDR
Keywords :
fast Fourier transforms; speech recognition; endpoint detection; energy difference degree; fast Fourier transform; formants; speech recognition; voyage data recorder; Acoustic noise; Acoustical engineering; Automation; Cepstral analysis; Educational institutions; Power engineering and energy; Signal analysis; Signal processing; Speech analysis; Speech recognition; EDD Parameter; Endpoint detection; Formants; Speech recognition; Voyage Data Recorder;
Conference_Titel :
Intelligent Control and Automation, 2006. WCICA 2006. The Sixth World Congress on
Conference_Location :
Dalian
Print_ISBN :
1-4244-0332-4
DOI :
10.1109/WCICA.2006.1713855
