ETRW on Speech Processing in Adverse Conditions Cannes-Mandelieu, France November 10-13, 1992

Hidden Markov Model Decomposition Recognition of Speech in Noise: A comprehensive experimental study

M. Kadirkamanathan

Speech Research Unit, Defence Research Agency, Malvern, England, UK

This paper accounts the performance of the Hidden Markov Model (HMM) decomposition recognition algorithm for speech recognition in a number of stationary and non-stationary background noises. Speaker dependent and speaker independent digit recognition tasks are evaluated and the results are also compared with those of a similar 'model adaptation to noise' scheme.

Full Paper

Bibliographic reference.  Kadirkamanathan, M. (1992): "Hidden Markov model decomposition recognition of speech in noise: a comprehensive experimental study", In SPAC-1992, 187-190.