Title :
Deleted smoothing of HMM parameters in speech recognition
Author :
Kim, N.S. ; Un, C.K.
Author_Institution :
Korea Adv. Inst. of Sci. & Technol., Taejon, South Korea
fDate :
4/29/1993 12:00:00 AM
Abstract :
A technique for smoothing hidden Markov model parameters based on the concepts of deleted estimation and probabilistic mapping is proposed. The proposed algorithm is closely related to deleted interpolation in its approach and is shown to yield higher recognition rate than the distance-based smoothing and co-occurrence smoothing methods.
Keywords :
hidden Markov models; interpolation; probability; speech recognition; HMM parameters; deleted estimation; deleted interpolation; deleted smoothing; hidden Markov model; probabilistic mapping; recognition rate; speech recognition;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:19930493
