지원사업
학술연구/단체지원/교육 등 연구자 활동을 지속하도록 DBpia가 지원하고 있어요.
커뮤니티
연구자들이 자신의 연구와 전문성을 널리 알리고, 새로운 협력의 기회를 만들 수 있는 네트워킹 공간이에요.
초록· 키워드
오류제보하기
In the Vector Taylor Series (VTS)-based noisy speech recognition methods, Hidden Markov Models (HMM) are usually trained with clean speech. However, better performance is expected by training the HMM with noisy speech. In a previous study, we could find that Minimum Mean Square Error (MMSE) estimation of the training noisy speech in the log-spectrum domain produce improved recognition results, but since the proposed algorithm was done in the log-spectrum domain, it could not be used for the HMM adaptation. In this paper, we modify the previous algorithm to derive a novel mathematical relation between test and training noisy speech in the cepstrum domain and the mean and covariance of the Multi-condition Training (MTR) trained noisy speech HMM are adapted. In the noisy speech recognition experiments on the Aurora 2 database, the proposed method produced 10.6% of relative improvement in Word Error Rates (WERs) over the MTR method while the previous MMSE estimation of the training noisy speech produced 4.3% of relative improvement, which shows the superiority of the proposed method.
목차
ABSTRACT
1. 서론
2. VTS 기반의 잡음음성 HMM 모델 적응
3. 실험결과
4. 결론
참고문헌
참고문헌 (10)
참고문헌 신청
1. [학위논문] - Gales, M. / 1995 / Model based techniques for noise- robust speech recognition / Ph.D. / University of Cambridge
2. [학술지(정기간행물)] - Ball, S. F / 1979 / Suppression of acoustic noise in speech using spectral subtraction / IEEE Trans. Acoust., Speech, Signal Process 27 (2) : 113 ~ 122
3. [학위논문] - Moreno, P. J. / 1996 / Speech recognition in noisy environments / Ph.D. / Carnegie Mellon University
4. [학술대회논문] - Hirsch, H. G. / 2000 / The Aurora experimental framework for the performance evaluation of speech recognition systems under noisy conditions / in Proc. ICSLP : 18 ~ 20
5. [학술지(정기간행물)] - Xu, H. / 2007 / Noise condition-dependent training based on noise classification and SNR estimation / IEEE Trans. Audio, Speech and Language Process 15 (8) : 2431 ~ 2443
2. [학술지(정기간행물)] - Ball, S. F / 1979 / Suppression of acoustic noise in speech using spectral subtraction / IEEE Trans. Acoust., Speech, Signal Process 27 (2) : 113 ~ 122
3. [학위논문] - Moreno, P. J. / 1996 / Speech recognition in noisy environments / Ph.D. / Carnegie Mellon University
4. [학술대회논문] - Hirsch, H. G. / 2000 / The Aurora experimental framework for the performance evaluation of speech recognition systems under noisy conditions / in Proc. ICSLP : 18 ~ 20
5. [학술지(정기간행물)] - Xu, H. / 2007 / Noise condition-dependent training based on noise classification and SNR estimation / IEEE Trans. Audio, Speech and Language Process 15 (8) : 2431 ~ 2443
이 논문의 저자 정보
이 논문과 함께 이용한 논문
파형 특징 추출과 신경망 학습 기반 모음 ‘ㅣ’ 음성 인식
노원빈
,
이종우
,
이재원
정보과학회 컴퓨팅의 실제 논문지
2016
.02
Reliability-weighted HMM Considering Inexact Observation For Enhancing Recognition Performance
Md. Ta
,
iquzzaman
,
Jin Young Kim
외 3명
한국정보기술학회논문지
2011
.02
가중치를 공유하는 신경회로망을 이용한 필기체 한글문자 인식 ( The Recognition of Handwritten Korean Characters by a Neural Network with Shared Weights )
전윤호
,
최종호
전자공학회논문지-B
1991
.08
분절모델 기반의 소리 이벤트 검출
김수종
,
정용주
한국정보기술학회논문지
2023
.06
최근 본 자료
댓글(0)
0
UCI(KEPA) : I410-ECN-0101-2015-700-001679281