DocumentCode :
312490
Title :
A new non-parametric estimation of Gabor spectrum envelopes
Author :
Miyanaga, Yoshikazu ; Imai, Hideaki ; Tochinai, Koji
Author_Institution :
Fac. of Eng., Hokkaido Univ., Sapporo, Japan
Volume :
1
fYear :
1996
fDate :
26-29 Nov 1996
Firstpage :
1
Abstract :
This paper introduces a new technique of spectrum envelope estimation which can be executed in a parallel/pipelined VLSI system. This method is applied to time-varying waveforms. In order to evaluate better spectrum envelopes from waves, FFT or Gabor transform is used at the first stage, and then a self-organized network with a supervisor is applied in the second stage
Keywords :
VLSI; fast Fourier transforms; learning (artificial intelligence); parallel architectures; parameter estimation; pipeline processing; self-organising feature maps; spectral analysis; speech processing; FFT; Gabor spectrum envelopes; Gabor transform; nonparametric estimation; parallel/pipelined VLSI system; self-organized network; spectrum envelopes; speech data; speech waveform; supervisor; time-varying waveforms; Convergence; Costs; Fires; Multi-layer neural network; Neural networks; Signal processing; Spectral analysis; Speech; Time varying systems; Very large scale integration;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
TENCON '96. Proceedings., 1996 IEEE TENCON. Digital Signal Processing Applications
Conference_Location :
Perth, WA
Print_ISBN :
0-7803-3679-8
Type :
conf
DOI :
10.1109/TENCON.1996.608683
Filename :
608683
Link To Document :
بازگشت