| H91-1008 | with the Lincoln Tied-Mixture | HMM CSR | System " ( in this Proceedings |
| H93-1094 | corpus , using a tied-mixture | HMM CSR | ap - proach . Recent work has |
| H89-1023 | word accuracy ( SI ) . The robust | HMM CSR | has Mso been integrated in real-time |
| H92-1077 | made to the Lincoln Tied-Mixture | HMM CSR | , to achieve effective operation |
| H92-1035 | phoneme insertion penalties in a | HMM CSR | system . SEGMENTAL NEURAL NET |
| H93-1094 | 20K , and up to 64K words . The | HMM CSR | has been converted to a stack-decoder-based |
| H91-1055 | time synchronous ( TS ) decoder | HMM CSR | \ -LSB- 18 , 20 , 21 \ -RSB- |
| H91-1010 | with the Lincoln Tied-Mixture | HMM CSR | System 1 </title> Douglas B Paul |
| H94-1102 | largevocabulary tasks , the Lincoln | HMM CSR | has been converted to use a stack |
| H92-1035 | performance of a state-of-the-art | HMM CSR | system . We have used the N-best |
| H92-1109 | and the resulting tied - mixture | HMM CSR | system achieved state-of-the-art |
| H94-1102 | database . More recently , the | HMM CSR | has been extended to tasks with |
| H91-1055 | of the Lincoln time-synchronous | HMM CSR | has been converted to a stack |
| H94-1102 | large-vocabulary stack decoder-based | HMM CSR | , with particular focus on improvement |
| H91-1091 | conversion of a version of the | HMM CSR | to use a stack decoder controlled |
| H92-1081 | <title> The Lincoln Large-Vocabulary | HMM CSR | * </title> Douglas B Paul ABSTRACT |
| H90-1089 | control structure , and convert the | HMM CSR | to this control structure ; ( |
| H89-2067 | development and integration into | HMM CSR | system of tied-mixture and word-context-dependent |
| H92-1081 | original Lincoln Tied-Mixture | HMM CSR | was implemented as a single-layer |
| H94-1079 | Large-Vocabulary Stack-Decoder Based | HMM CSR | * </title> Douglas B Paul Abstract |
