Opens in a new tab

AES E-Library

← Back to search

Engineering Brief

Examining Influence of Distance to Microphone on Accuracy of Speech Recognition

Authors: Bratoszewski, Piotr; Szykulski, Marcin; Czyzewski, Andrzej

AES Convention 138 · Paper 194 · May 2015

Abstract

The problem of controlling a machine by the distant-talking speaker without a necessity of handheld or body-worn equipment usage is considered. A laboratory setup is introduced for examination of performance of the developed automatic speech recognition system fed by direct and by distant speech acquired by microphones placed at three different distances from the speaker (0.5 m to 1.5 m). For feature extraction from the voice signal the Mel- Frequency Cepstral Coefficients (MFCC) are used. The experiments are conducted employing the HTK engine (Hidden Markov Toolkit) for the Automatic Speech Recognition (ASR) task. The dictionary of 184 words was employed and WER (Word Error Rate), correctness and accuracy measures were calculated in order to verify and to compare obtained results of speech recognition.

Details

Published in
AES Convention 138
AES Convention
138
Paper number
194
Publication date
May 6, 2015
Affiliation
Gdansk University of Technology, Gdansk, Poland (See document for exact affiliation information.)
Type
Engineering Brief