| H92-1037 | speaker variation effects for | speakerindependent speech recognition | . The network is used as a nonlinear |
| H92-1034 | detailed pronunciation variations for | speakerindependent speech recognition | . 4 NEW WORD LEARNING In dictation |
| A97-1005 | recognizers use an HMM-based continuous | speakerindependent speech recognition | technology for PC 's under Windows |
| P15-1105 | al. , 2000 ) primarily because | speakerindependent speech recognition | is a tough computer science problem |
| H91-1054 | also enhance the robustness of | speakerindependent speech recognition | . Preliminary results indicate |
| H91-1054 | also enhance the robustness of | speakerindependent speech recognition | \ -LSB- 10 \ -RSB- . Normalization |
| H89-2037 | has shown that highly accurate | speakerindependent speech recognition | is possible , thus alleviating |
| H91-1054 | as part of the front-end of the | speakerindependent speech recognition | system as shown in Figure 2 . |
| H91-1054 | . Sphinx , a state-of-the-art | speakerindependent speech recognition | system developed at CMU \ -LSB- |
| A00-1046 | Application Factory , a continuous , | speakerindependent speech recognition | system , was used with a bigram |
