Spoken Sentence Recognition Based on HMM-LR with Hybrid Language Modeling (Special Issue on Natural Language Processing and Understanding)
スポンサーリンク
概要
- 論文の詳細を見る
This paper describes Japanese spoken sentence recognition using hybrid language modeling, which combines the advantages of both syntactic and stochastic language models. As the baseline system, we adopted the HMM-LR speech recognition system, with which we have already achieved good performance for Japanese phrase recognition tasks. Several improvements have been made to this system aimed at handling continuously spoken sentences. The first improvement is HMM training with continuous utterances as well as word utterances. In previous implementations, HMMs were trained with only word utterances. Continuous utterances are included in the HMM training data because coarticulation effects are much stronger in continuous utterances. The second improvement is the development of a sentential grammar for Japanese. The sentential grammar was created by combining inter-and intra-phrase CFG grammars, which were developed separately. The third improvement is the incorporation of stochastic linguistic knowledge, which includes stochastic CFG and a bigram model of production rules. The system was evaluated using continuously spoken sentences from a conference registration task that included approximately 750 words. We attained a sentence accuracy of 83.9% in the speaker-dependent condition.
- 社団法人電子情報通信学会の論文
- 1994-02-25
著者
-
Kita Kenji
Faculty of Engineering, Tokushima University
-
Morimoto Tsuyoshi
ATR Interpreting Telecommunications Research Laboratories
-
Sagayama S
Atr Interpreting Telephony Research Lab.
-
Sagayama Shigeki
Ntt Human Interface Laboratories
-
Ohkura Kazumi
Information & Communication systems Research Center, SANYO Electric Co., Ltd.
-
Yano Yaneo
Faculty of Engineering, Tokushima University
-
Kita K
Faculty Of Engineering Tokushima University
-
Kita Kenji
Faculty Of Engineering Tokushima University
-
Yano Yaneo
Faculty Of Engineering Tokushima University
-
Morimoto T
Atr Interpreting Telecommunications Res. Lab. Kyoto Jpn
-
Ohkura Kazumi
Information & Communication Systems Research Center Sanyo Electric Co. Ltd.
-
Kita Kenji
Faculty Of Engineering The University Of Tokushima
関連論文
- One-Pass Search Algorithm for Continuous Speech Recognition Using Generalized LR Parsing : A CFG-Driven, Frame-Synchronous HMM-Based Approach
- A voice conversion based on phoneme segment mapping
- Spoken Sentence Recognition Based on HMM-LR with Hybrid Language Modeling (Special Issue on Natural Language Processing and Understanding)
- LR Parsing with a Category Reachability Test Applied to Speech Recognition (Special Issue on Speech and Discourse Processing in Dialogue Systems)
- Semantic-level Transfer in Japanese-German Speech Translation : Some Experiences
- FUSING MULTIMODAL INPUTS
- ANALYSIS AND INTEGRATION OF MULTIMODAL INPUTS IN INTERPRETING TELECOMMUNICATIONS
- パターンを用いた曖昧性の検出と再確認方式
- Speaker-Consistent Parsing for Speaker-Independent Continuous Speech Recognition
- A Unification-Based Japanese Parser for Speech-to-Speech Translation (Special Issue on Speech and Discourse Processing in Dialogue Systems)
- Analysis of Gestures in a Multimedia/Multimodal Interpreting Experiment
- Automatic Determination of the Number of Mixture Components for Continuous HMMs Based on a Uniform Variance Criterion
- An Empirical Study on Rule Granularity and Unification Interleaving in Unification-Based Parsers
- Unsupervised Speaker Adaptation Using All-Phoneme Ergodic Hidden Markov Network
- Continuous Speech Recognition Using a Combination of Syntactic Constraints and Dependency Relationships
- An lnformation-Theoretic Model of Discourse for Next Utterance Type Prediction
- Speech Recognition Using Function-Word N-Grams and Content-Word N-Grams
- Skeleton Pruning Based on the Total Bisector Angle of the End Branches
- Discriminative Training Based on Minimum Classification Error for a Small Amount of Data Enhanced by Vector-Field-Smoothed Bayesian Learning