Sixth International Conference on Spoken Language Processing
(ICSLP 2000)

Beijing, China
October 16-20, 2000

Speaker Dependent Temporal Constraints Combined with Speaker Independent HMM for Speech Recognition in Noise

NÚstor Becerra Yoma

Dept. of Electrical Engineering/University of Chile, Santiago, Chile

This paper addresses the problem of speech recognition in noise using speaker-dependent temporal constraints in the Viterbi algorithm in combination with speaker-independent HMM. It is shown that the speaker-dependent re-estimation of state duration parameters requires a low computational load and a small training database, and can lead to reductions in the error rate as high as 30% or 40% with clean signals and with signals corrupted by additive noise, without noise canceling methods. Moreover, the approach here covered could also be seen as a speaker adaptation method in which only temporal restrictions parameters are adapted.


Full Paper

Bibliographic reference.  Yoma, NÚstor Becerra (2000): "Speaker dependent temporal constraints combined with speaker independent HMM for speech recognition in noise", In ICSLP-2000, vol.3, 526-529.