Faculty Advisor

Hakim, Hossein

Abstract

The goal of this study was to design a system that would facilitate word processing for the handicapped, and implement the general public's most natural and effective way of communication--verbally. The word processor recognized the letters of the alphabet. This project explored and tested three methods of feature extraction which were the DFT, Spectrogram, and MFCCs. It was shown that the MFCC provides the best feature vector with a recognition rate of 97.9%.

Publisher

Worcester Polytechnic Institute

Date Accepted

January 2002

Major

Electrical Engineering

Project Type

Major Qualifying Project

Accessibility

Restricted-WPI community only

Advisor Department

Electrical and Computer Engineering

Advisor Program

Electrical and Computer Engineering

Share

COinS