COST278 and ISCA Tutorial and Research Workshop (ITRW) on Robustness Issues in Conversational Interaction University of East Anglia, Norwich, UK August 30-31, 2004

Hidden Markov Models in Voice Activity Detection

Jirí Tatarinov, Petr Pollák

Czech Technical University, Faculty of Electrical Engineering, CTU FEE K13131, Prague, Czech Republic

This paper describes two algorithms for speech/pause detection based on Hidden Markov Models. There are proposed algorithm based on separate training and testing and algorithm on simultaneous training-testing procedure. Algorithm are compared to the differential cepstral detector. HMM based algorithms are successful especially under low SNR . Under higher SNR they reach comparable results to the cepstral detector.

Full Paper

Bibliographic reference.  Tatarinov, Jirí / Pollák, Petr (2004): "Hidden Markov models in voice activity detection", In Robust2004, paper 37.