| N01-1017 | matching segments were used for | acoustic adaptation | training . The SI model had been |
| N01-1017 | not an appropriate domain for | acoustic adaptation | methods . However , adapted language |
| H91-1004 | should be faster than in normal | acoustic adaptation | . Modelling the source of variation |
| W15-5121 | off-line fMLLR linear transformation | acoustic adaptation | was performed . The acoustic |
| N01-1017 | recognition , evaluating accuracy and | acoustic adaptation | . ( Valtchev , et al. ( 1998 |
| W13-3903 | dependence on ASR system Despite the | acoustic adaptation | , there is a great variability |
| N01-1017 | transcriptions can be successfully used in | acoustic adaptation | . However , non-literal transcriptions |
| W13-3903 | level of autonomy Despite the | acoustic adaptation | , there might be a disparity |
| N01-1017 | transcriptions was evaluated in two ways : | acoustic adaptation | and language modeling . 4.1 Adapted |
| W10-4307 | speech recognition . Per-user | acoustic adaptation | is used to improve recognition |
| N13-2005 | language identification and used | acoustic adaptation | to improve it . Choueiter et |
| D09-1118 | recognition passes , including | acoustic adaptation | and language model rescoring |
| N04-4007 | Vocal Tract Normalization and | Acoustic Adaptation | : We further extend on our baseline |
