Title :
Nonspecific speech recognition method based on composite LVQ1 and LVQ2 network
Author :
Liang, Shuling ; Wang, Chaoli ; Du, Jiaming
Author_Institution :
Sch. of Opt.-Electr. & Comput. Eng., Univ. of Shanghai for Sci. & Technol., Shanghai, China
Abstract :
A novel method of normalization is proposed in this paper, in which the MFCC (Mel frequency Cepstral Coefficient) and DeltaMFCC (difference mel frequency cepstral coefficient) are sampled equidistantly. For these normalized signals, a new speech recognition based on composite LVQ1(Learning Vector Quantization) network and LVQ2(Improved Learning Vector Quantization) network is presented. First, MFCC and DeltaMFCC feature extraction algorithms are introduced, then their coefficients are normalized. The recognition is first to learn coarsely by LVQ1 network and then to learn finely by LVQ2 network. Finally the simulation is given, which shows that the proposed algorithm improves the recognition rates effectively, with shorter training time in comparison with LVQ1 network used alone.
Keywords :
cepstral analysis; feature extraction; speech recognition; vector quantisation; DeltaMFCC; MFCC; composite LVQ1 network; composite LVQ2 network; difference mel frequency cepstral coefficient; feature extraction algorithm; improved learning vector quantization network; learning vector quantization network; mel frequency cepstral coefficient; nonspecific speech recognition method; Cepstral analysis; Equations; Feature extraction; Hidden Markov models; Humans; Mel frequency cepstral coefficient; Optical computing; Signal processing algorithms; Speech recognition; Vector quantization; Learning Vector Quantization; Mel Frequency Cepstral Coefficient; Neural Network; Speech Recognition;
Conference_Titel :
Control and Decision Conference, 2009. CCDC '09. Chinese
Conference_Location :
Guilin
Print_ISBN :
978-1-4244-2722-2
Electronic_ISBN :
978-1-4244-2723-9
DOI :
10.1109/CCDC.2009.5192778
