Connected word recognition in extreme noisy environment using weighted state probabilities (wsp)

T. Vaich; A. Cohen

Connected word recognition in extreme noisy environment using weighted state probabilities (wsp)

T. Vaich, A. Cohen

Department of Electrical & Computer Engineering

Research output: Contribution to journal › Conference article › peer-review

Abstract

Recognition of continuous speech in extreme noisy environments is a difficult task. A novel algorithm is suggested to enhance the performance of recognition in very low SNRs. The left to right HMM Weighted State Probabilities (WSP) method considers not only the probability of getting the given observation sequence, but also the pattern of states probabilities. On a ten digits (Hebrew) recognition task, with SNR of 10 db, the WSP has improved recognition results from 0% to 50%. It is suggested to apply the method, in conjunction with PMC enhancement algorithm, to very low SNR word spotting systems.

Original language	English
Journal	European Signal Processing Conference
State	Published - 1 Jan 2015
Event	8th European Signal Processing Conference, EUSIPCO 1996 - Trieste, Italy Duration: 10 Sep 1996 → 13 Sep 1996

ASJC Scopus subject areas

Signal Processing
Electrical and Electronic Engineering

Cite this

@article{4ad48e05cb3e488f918d2fd865cde713,

title = "Connected word recognition in extreme noisy environment using weighted state probabilities (wsp)",

abstract = "Recognition of continuous speech in extreme noisy environments is a difficult task. A novel algorithm is suggested to enhance the performance of recognition in very low SNRs. The left to right HMM Weighted State Probabilities (WSP) method considers not only the probability of getting the given observation sequence, but also the pattern of states probabilities. On a ten digits (Hebrew) recognition task, with SNR of 10 db, the WSP has improved recognition results from 0% to 50%. It is suggested to apply the method, in conjunction with PMC enhancement algorithm, to very low SNR word spotting systems.",

author = "T. Vaich and A. Cohen",

note = "Publisher Copyright: {\textcopyright} 2015 European Signal Processing Conference, EUSIPCO. All rights reserved.; 8th European Signal Processing Conference, EUSIPCO 1996 ; Conference date: 10-09-1996 Through 13-09-1996",

year = "2015",

month = jan,

day = "1",

language = "English",

journal = "European Signal Processing Conference",

issn = "2219-5491",

}

TY - JOUR

T1 - Connected word recognition in extreme noisy environment using weighted state probabilities (wsp)

AU - Vaich, T.

AU - Cohen, A.

PY - 2015/1/1

Y1 - 2015/1/1

N2 - Recognition of continuous speech in extreme noisy environments is a difficult task. A novel algorithm is suggested to enhance the performance of recognition in very low SNRs. The left to right HMM Weighted State Probabilities (WSP) method considers not only the probability of getting the given observation sequence, but also the pattern of states probabilities. On a ten digits (Hebrew) recognition task, with SNR of 10 db, the WSP has improved recognition results from 0% to 50%. It is suggested to apply the method, in conjunction with PMC enhancement algorithm, to very low SNR word spotting systems.

AB - Recognition of continuous speech in extreme noisy environments is a difficult task. A novel algorithm is suggested to enhance the performance of recognition in very low SNRs. The left to right HMM Weighted State Probabilities (WSP) method considers not only the probability of getting the given observation sequence, but also the pattern of states probabilities. On a ten digits (Hebrew) recognition task, with SNR of 10 db, the WSP has improved recognition results from 0% to 50%. It is suggested to apply the method, in conjunction with PMC enhancement algorithm, to very low SNR word spotting systems.

UR - http://www.scopus.com/inward/record.url?scp=85113829801&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:85113829801

SN - 2219-5491

JO - European Signal Processing Conference

JF - European Signal Processing Conference

T2 - 8th European Signal Processing Conference, EUSIPCO 1996

Y2 - 10 September 1996 through 13 September 1996

ER -

Connected word recognition in extreme noisy environment using weighted state probabilities (wsp)

Abstract

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this