Comparing Jacobian adaptation with cepstral mean normalization and parallel model combination for noise robust speech recognition

ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

Volumn 1, Issue , 2002, Pages 193-196

  Pärssinen, Kimmo ^a   Salmela, Petri ^a   Harju, Mikko ^a   Kiss, Imre ^b  

^a TAMPERE UNIVERSITY OF TECHNOLOGY   (Finland)

^b NOKIA RESEARCH CENTER   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

ADDITIVE NOISE; CEPSTRAL MEAN NORMALIZATION; JACOBIAN ADAPTATION; PARALLEL MODEL COMBINATION;

CALCULATIONS; COSINE TRANSFORMS; DATABASE SYSTEMS; MATRIX ALGEBRA; SIGNAL TO NOISE RATIO; SPEECH ANALYSIS; SPURIOUS SIGNAL NOISE; SYSTEMS ANALYSIS;

SPEECH RECOGNITION;

EID: 0036293691     PISSN: 15206149     EISSN: None     Source Type: Journal    
DOI: 10.1109/ICASSP.2002.5743687     Document Type: Article

References (9)

1. Flexible speech recognition
2. Environmental adaptation based on first order approximation
3. Jacobian approach to fast acoustic model adaptation
4. Effectiveness of linear prediction characteristics of speech wave automatic speaker identification and verification
5. Model-based techniques for noise robust speech recognition
6. Improved Jacobian adaptation for fast acoustic model adaptation in noisy speech recognition
7. Improving environmental robustness in large vocabulary speech recognition
8. Maximum likelihood from incomplete data via the EM algorithm