University of Bologna

Dipartimento di Elettronica, Informatica e Sistemistica

Publication Database Query

Searched for "esscirc97:Fer97"
Clicking on the number of each publication you can retrieve the bibtex entry. If available, the abstract can be accessed clicking on the title of the work.
(Conference)

A 0.9V, 30uW Feature Extractor for Remote Speech Recognition

A. Ferrari and M. Borgatti and M. Felici and R. Guerrieri

Abstract:

A feature-extraction chip for speech recognition computes fifteen cepstra each 8ms at 64kHz clock rate and dissipates 30uW at 0.9V. It has been implemented as a gate array in a 0.5um, three-metal CMOS technology. The average energy required to process a single word of the TI46 speech corpora is 10uJ. It achieves recognition rates over 98% in isolated-word, speech recognition tasks. The main contribution of this paper is the description of the implementation of an integrated circuit that implements a novel, highly simplified algorithm achieving efficient speech feature computation and compression, and very low-power and low-voltage operation.

Microelectronics Group Home Page Staff
Research Activities Teaching Activity
Publications Local Stuff
DEIS Home Page University of Bologna Home Page

External link, Local link, For local sites only, Stiil empty.