DocumentCode :
1652896
Title :
Robust speech recognition using noise-cluster HMM interpolation
Author :
Thatphithakkul, Nattanun ; Kruatrachue, Boontee ; Wutiwiwatchai, Chai ; Marukatat, Sanparith ; Boonpiam, Vataya
Author_Institution :
Comput. Eng. Dept., King Mongkut´´s Inst. of Technol. Ladkrabang, Bangkok
fYear :
2008
Firstpage :
596
Lastpage :
600
Abstract :
This paper proposes a novel approach called noise-cluster HMM interpolation for robust speech recognition. The approach helps alleviating the problem of speech recognition under noisy environments not trained in the system. In this method, a new HMM is interpolated from existing noisy-speech HMMs that are best matched to the input speech. This process is performed on-the-fly with an acceptable delay time and, hence, no need to prepare and store the final model in advance. Interpolation weights among HMMs can be determined by either a direct or a tree-structured search. Evaluated focusing on speech in unseen noisy-environments, the proposed method obviously outperforms a baseline system whose acoustic model for such unseen environment is selected from a tree structure.
Keywords :
hidden Markov models; interpolation; speech recognition; tree searching; acoustic model; hidden Markov models; noise-cluster HMM interpolation; speech recognition; time delay; tree-structured search; Acoustic noise; Delay effects; Hidden Markov models; Impedance matching; Interpolation; Noise robustness; Speech analysis; Speech recognition; Tree data structures; Working environment noise;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Signal Processing, 2008. ICSP 2008. 9th International Conference on
Conference_Location :
Beijing
Print_ISBN :
978-1-4244-2178-7
Electronic_ISBN :
978-1-4244-2179-4
Type :
conf
DOI :
10.1109/ICOSP.2008.4697203
Filename :
4697203
Link To Document :
بازگشت