2005 IEEE International Conference on Acoustics, Speech, and Signal Processing
March 18-23, 2005 • Pennsylvania Convention Center/Marriott Hotel • Philadelphia, PA, USA
ICASSP
30th Anniversary
SP-L5: Discriminative Training |
| Session Type: Lecture |
| Time: Monday, March 21, 15:30 - 17:30 |
| Location: CC: Room 113-A |
| Chair: Roland Kuhn, National Research Council |
| SP-L5.1: ADAPTATION OF PRECISION MATRIX MODELS ON LARGE VOCABULARY CONTINUOUS SPEECH RECOGNITION |
| Khe Chai Sim; Cambridge University |
| Mark J. F. Gales; Cambridge University |
| SP-L5.2: DISCRIMINATIVE TRAINING OF CDHMMS FOR MAXIMUM RELATIVE SEPARATION MARGIN |
| Chaojun Liu; York University |
| Hui Jiang; York University |
| Xinwei Li; York University |
| SP-L5.3: STATISTICAL PERFORMANCE ANALYSIS OF MCE/GPD LEARNING IN GAUSSIAN CLASSIFIERS AND HIDDEN MARKOV MODELS |
| Mohamed Afify; BBN Technologies |
| Xin-Wei Li; York University |
| Hui Jiang; York University |
| SP-L5.4: DISCRIMINATIVE TRAINING OF ACOUSTIC MODELS APPLIED TO DOMAINS WITH UNRELIABLE TRANSCRIPTS |
| Lambert Mathias; Johns Hopkins University |
| Girija Yegnanarayanan; Multimodal Technologies, Inc. |
| Juergen Fritsch; Multimodal Technologies, Inc. |
| SP-L5.5: MINIMUM CLASSIFICATION ERROR FOR LARGE SCALE SPEECH RECOGNITION TASKS USING WEIGHTED FINITE STATE TRANSDUCERS |
| Erik McDermott; NTT Corporation |
| Shigeru Katagiri; NTT Corporation |
| SP-L5.6: DISCRIMINATIVE TRAINING BASED ON THE CRITERION OF LEAST PHONE COMPETING TOKENS FOR LARGE VOCABULARY SPEECH RECOGNITION |
| Bo Liu; University of Sci. & Tech. of China |
| Hui Jiang; York University |
| Jian-Lai Zhou; Microsoft Reseach Asia |
| Ren-Hua Wang; University of Sci. & Tech. of China |