Opens in a new tab

AES E-Library

← Back to search

Journal Article

A Procedure for the Segmentation of Connected Speech

Authors: Reddy, D. Raj; Vicens, Pierre J.

Journal of the Audio Engineering Society · Volume 16 · Issue 4 · pp. 404–411 · October 1968

Abstract

Smoothing and differencing operations on the digitized acoustic waveform generate parameters which are used to determine whether the characteristics of a sound are changing or similar; parts that possess similar parameters are grouped together to form sustained segments and parts that possess changing parameters form transitional segments, resulting in the segmentation of connected speech into parts approximately corresponding to phonemes.

Details

Publication
Journal of the Audio Engineering Society
Volume
16
Issue
4
Pages
404–411
Publication date
October 6, 1968
Affiliation
Computer Science Department, Stanford University, Stanford, CA (See document for exact affiliation information.)
Type
Journal Article