Incorporation of time varying ar modeling in speech recognition system based on dynamic programming

Zafrir Babin, Felix A. Flomen, Ilan D. Shallom

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

In this paper, two isolated-word speaker-independent recognition systems are presented. These systems are based on two non-stationary representations of the speech signal first being applied in speech recognition application. This non stationary representation is utilized to make direct use of the speech dynamic features. The first system is based on a Time-Varying Linear Prediction model (TVLP). This recognition system is distinguishable from all others by its incorporation of the model and appropriate extension of the well-known Log Likelihood Ratio (LLR) distortion measure. The second system is based on the lattice model which is represented by Time-Varying Reflection Coefficients (TVRC). In this system, the distortion measure used is the Euclidian distance. In order to properly support this system, the abovementioned extension of the LLR was modified for use with the time-varying reflection coefficients.

Original languageEnglish
Title of host publicationProceedings - 17th Convention of Electrical and Electronics Engineers in Israel, EEIS 1991
PublisherInstitute of Electrical and Electronics Engineers
Pages289-292
Number of pages4
ISBN (Electronic)0879426780, 9780879426781
DOIs
StatePublished - 1 Jan 1991
Event17th Convention of Electrical and Electronics Engineers in Israel, EEIS 1991 - Tel Aviv, Israel
Duration: 5 Mar 19917 Mar 1991

Publication series

NameProceedings - 17th Convention of Electrical and Electronics Engineers in Israel, EEIS 1991

Conference

Conference17th Convention of Electrical and Electronics Engineers in Israel, EEIS 1991
Country/TerritoryIsrael
CityTel Aviv
Period5/03/917/03/91

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Incorporation of time varying ar modeling in speech recognition system based on dynamic programming'. Together they form a unique fingerprint.

Cite this