Title :
On modeling duration in context in speech recognition
Author_Institution :
Texas Instrum. Inc, Dallas, TX, USA
Abstract :
A clustering algorithm is introduced that allows clustering of HMM (hidden Markov models) models directly. This clustering algorithm determines the appropriate duration profile for a recognition unit. High-performance speaker-independent digit recognition on a studio-quality connected-digit database is demonstrated using this algorithm
Keywords :
Markov processes; speech recognition; HMM model; clustering algorithm; contextual effects; duration profile; hidden Markov models; seed models; speaker-independent digit recognition; speech recognition; studio-quality connected-digit database; Clustering algorithms; Context modeling; Degradation; Hidden Markov models; Instruments; Laboratories; Power system modeling; Spatial databases; Speech recognition; Training data;
Conference_Titel :
Acoustics, Speech, and Signal Processing, 1989. ICASSP-89., 1989 International Conference on
Conference_Location :
Glasgow
DOI :
10.1109/ICASSP.1989.266455
