A low-power VLSI feature extractor for speech recognition
Authors:
- M. BorgattiMBM. Borgatti
- A. FerrariAFA. Ferrari
- R. GuerrieriRGR. Guerrieri
- 27 November 2002
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 5 (15206149) , 3061-3064
Abstract
A low-power feature extraction chip computing cepstral coefficients from linear predictive analysis on one-bit quantized speech signals is presented and its VLSI implementation is evaluated. An isolated-word small-vocabulary speech recognizer based on these features has been developed. Its recognition accuracy is within 2% below a system based on standard linear predictive cepstral features. The power consumption of the feature extractor chip is 30 /spl mu/W at 0.9 V.Keywords
This publication has 5 references indexed in Scilit:
- Signal modeling techniques in speech recognitionProceedings of the IEEE, 1993
- Low-power CMOS digital designIEEE Journal of Solid-State Circuits, 1992
- Zero-crossing-based linear prediction for speech recognitionElectronics Letters, 1989
- Linear prediction: A tutorial reviewProceedings of the IEEE, 1975
- Effects of Amplitude Distortion upon the Intelligibility of SpeechThe Journal of the Acoustical Society of America, 1946