Title :
Classified nonlinear predictive vector quantization of speech spectral parameters
Author :
Loo, James H Y ; Chan, Wad-Yip ; Kabal, Peter
Author_Institution :
Dept. of Electr. Eng., McGill Univ., Montreal, Que., Canada
Abstract :
Nonlinear predictive split vector quantization (NPSVQ) and classified NPSVQ (CNPSVQ) are introduced to exploit the correlation among the speech spectral parameters from two adjacent analysis frames. By interleaving intraframe SVQ with forward predictive SVQ, the error propagation is limited to at most one adjacent frame. At an overall bit rate of about 21 bits/frame, NPSVQ can provide similar coding quality as intraframe SVQ at 24 bits/frame. Voicing classification as used in CNPSVQ to obtain an additional average gain of 1 bit/frame for unvoiced frames. Therefore, an overall bit rate of 20 bits/frame is obtained for unvoiced frames. The particular form of nonlinear prediction we use incurs virtually no additional encoding computational complexity. We have verified our comparative performance results using subjective listening tests
Keywords :
correlation methods; prediction theory; spectral analysis; speech coding; vector quantisation; CNPSVQ; analysis frames; average gain; classified NPSVQ; classified nonlinear predictive vector quantization; coding quality; correlation; error propagation; forward predictive SVQ; intraframe SVQ; overall bit rate; performance results; speech spectral parameters; subjective listening tests; unvoiced frames; voicing classification; Bit rate; Computer errors; Encoding; Filters; Interleaved codes; Speech analysis; Speech coding; Speech processing; Speech synthesis; Vector quantization;
Conference_Titel :
Acoustics, Speech, and Signal Processing, 1996. ICASSP-96. Conference Proceedings., 1996 IEEE International Conference on
Print_ISBN :
0-7803-3192-3
DOI :
10.1109/ICASSP.1996.543232
