Title :
Integrated polispectrum on speech recognition
Author :
Moreno, Asunciòn ; Rutllán, Miquel
Author_Institution :
Dept. de Teoria del Senyal i Comunicacions, Univ. Politecnica de Catalunya, Barcelona, Spain
Abstract :
The paper deals with the application of higher order statistics (HOS) in a recognition system where the speech input is corrupted by noise. Cumulants, bispectrum and trispectrum are being used in noisy speech signal processing because of their immunity to noise. The authors introduce the use of integrated polispectrum and trispectrum in a recognition system based on a filter bank analysis. The results improve other HOS based methods without a substantial increase of the computational load
Keywords :
acoustic filters; higher order statistics; spectral analysis; speech recognition; bispectrum; computational load; cumulants; filter bank analysis; higher order statistics; integrated polispectrum; integrated trispectrum; noise immunity; noise-corrupted speech input; noisy speech signal processing; speech recognition; trispectrum; Filter bank; Fourier transforms; Frequency estimation; Gaussian noise; Higher order statistics; Physics computing; Speech analysis; Speech enhancement; Speech processing; Speech recognition;
Conference_Titel :
Spoken Language, 1996. ICSLP 96. Proceedings., Fourth International Conference on
Conference_Location :
Philadelphia, PA
Print_ISBN :
0-7803-3555-4
DOI :
10.1109/ICSLP.1996.607844
