A New HMnet Construction Algorithm Requiring No Contextual Factors
スポンサーリンク
概要
- 論文の詳細を見る
Many methods have been proposed for constructing context-dependent phoneme models using Hidden Markov Models (HMMs) to improve performance. These conventional methods require previously defined contextual factors. If these factors are deficient, the method exhibit poor recognition performance. In this paper, we propose a new construction algorithm for HMnet which does not require pre-defined contextual factors. Experiments demonstrated that the new algorithm could construct the HMnet even for the case that the Successive State Splitting (SSS) algorithm could not. The new algorithm produced better phoneme recognition characteristics than the SSS algorithm.
- 社団法人電子情報通信学会の論文
- 1995-06-25
著者
-
Ito A
Graduate School Of Engineering Tohoku University
-
Akinori Ito
Graduate School Of Engineering Tohoku University
-
Aso Hirotomo
Faculty Of Engineering Tohoku University
-
Makino S
Graduate School Of Engineering Tohoku University
-
Makino Shozo
Graduate School Of Engineering Tohoku University
-
Ito Akinori
Graduate School Of Engineering Tohoku University
-
Suzuki Motoyuki
Institute Of Technology And Science The University Of Tokushima
-
Suzuki Motoyuki
Faculty of Engineering, Tohoku University
-
Makino Shozo
School of Information Science, Tohoku University
-
Ito Akinori
School of Information Science, Tohoku University
-
Shimodaira Hiroshi
School of Information science, JAIST
-
Shimodaira H
Bunkyo Univ. Chigasaki‐shi Jpn
-
Shimodaira Hiroshi
School Of Information Science Japan Advanced Institute Of Science And Technology
-
Suzuki Motoyuki
Faculty Of Engineering The University Of Tokushima
-
伊藤 彰則
Graduate School of Engineering, Tohoku University
-
ASO Hirotomo
Faculty of Engineering, Tohoku University
関連論文
- SIG-SLP/SIG-NL合同セッションここまでできるぞ音声/言語処理技術 : 音声編
- ここまでできるぞ音声/言語処理技術 : 音声編
- 連続音声認識コンソーシアム2002年度版ソフトウエアの概要
- 連続音声認識コンソーシアム2001年度版ソフトウエアの概要
- 日本語ディクテーション基本ソフトウェア(99年度版)
- 2000-NL-137-7 / 2000-SLP-31-2 日本語ディクテーション基本ソフトウェア(99年度版)の性能評価
- 2000-NL-137-7 / 2000-SLP-31-2 日本語ディクテーション基本ソフトウェア(99年度版)の性能評価
- 日本語ディクテーション基本ソフトウェア : 97年度版
- 日本語ディクテーション基本ソフトウェア(97年度版)
- 日本語ディクテーション基本ソフトウェア(97年度版)の性能評価
- 連続音声認識コンソーシアム2000年度版ソフトウエアの概要と評価
- 新博士によるパネルディスカッションIII 「私のための研究・価値を生み出す研究」
- 大語彙日本語連続音声認識研究基盤の整備 : 汎用音素モデルの作成
- 大語彙日本語連続音声認識研究基盤の整備 : 学習・評価テキストコーパスの作成
- 大語彙日本語連続音声認識研究基盤の整備 : 評価用連続音声認識プログラムの開発
- 「人はなぜコンピューターを人間として扱うか『メディアの等式』の心理学」, バイロン・リーブズ, クリフォード・ナス著, 細馬宏通訳, 翔泳社, 2001年(私のすすめるこの一冊,コーヒーブレーク)
- 日本語ディクテーション基本ソフトウェア(97年度版)の性能評価
- Stability Analysis of Continuous Culture in Diauxic Growth
- Development of a toxicity evaluation system for gaseous compounds using air-liquid interface culture of a human bronchial epithelial cell line, Calu-3
- Development of a Simple Double-layered Cell Culture System Using Caco-2 and TIG-1 Cells as a New Cytotoxicity Test
- A New Assay for Evaluating Hepatotoxicity and Cytotoxicity Using LDL-Uptake Activity of Liver Cells
- Rapid and Sensitive Neurotoxicity Test Based on the Morphological Changes of PC12 Cells with Simple Computer-Assisted Image Analysis
- ATTACHMENT KINETICS OF ANIMAL CELLS IMMEDIATELY AFTER CONTACT ONTO SPECIFIC AND/OR NON-SPECIFIC SURFACES
- Improved Reference Speaker Weighting Using Aspect Model
- Bit rate reduction of mixed excitation linear prediction coder by Lempel-Ziv segment quantization
- Selection of Optimum Vocabulary and Dialog Strategy for Noise-Robust Spoken Dialog Systems
- Pronunciation error detection for computer-assisted language learning system based on error rule clustering using a decision tree
- An Evaluation Method of Japanese Pronunciation for Korean Native Speakers
- I-069 Smile and Laugh Recognition from Natural Conversation Video
- A New HMnet Construction Algorithm Requiring No Contextual Factors
- Information Hiding for G.711 Speech Based on Substitution of Least Significant Bits and Estimation of Tolerable Distortion
- Source-filter separation for nonstationary voiced speech based on sinusoidal representation
- Fast optimization of language model weight and insertion penalty from n-best candidates
- Speech Recognition under Multiple Noise Environment Based on Multi-Mixture HMM and Weight Optimization by the Aspect Model
- The Performance Prediction on Sentence Recognition Using a Finite State Word Automaton
- Novel Tonal Feature and Statistical User Modeling for Query-by-Humming
- An Efficient Parallel Algorithm for the Solution of Block Tridiagonal Linear Systems
- Recognition of Devanagari Characters Using Neural Networks (Special Issue on Character Recognition and Document Understanding)
- A Highly Parallel Systolic Tridiagonal Solver
- Characteristics of landslide and failure with their Countermeasures in Kanmon-group Slopes, Yamaguchi prefecture, Western Japan
- A grammatical error detection method for dialogue-based CALL system
- 5 What Can be Done for Cardiovascular Medicine Using Robotics and Information Technologies?(Robotics and Information Technologies (IT) in the Field of Cardiovascular Medicine,Plenary Session 6 (PL6) (H),The 70th Anniversary Annual Scientific Meeting of th
- Automatic Determination Algorithm for Optimum Number of States in Discrete-Type HMnet
- Skeleton Pruning Based on the Total Bisector Angle of the End Branches
- Robust Transmission of Audio Signals over the Internet: An Advanced Packet Loss Concealment for MP3-Based Audio Signals
- Novel Tonal Feature and Statistical User Modeling for Query-by-Humming
- Novel Tonal Feature and Statistical User Modeling for Query-by-Humming
- 音声言語情報処理研究会の20年-歴代主査による研究レビュー-
- Foreword to the special issue on ``the speech communication and its related technologies''
- Robust Transmission of Audio Signals over the Internet: An Advanced Packet Loss Concealment for MP3-Based Audio Signals ( Fundamental Aspects and Recent Developments in Multimedia and VLSI Systems)