• Search Research Projects
  • Search Researchers
  • How to Use
  1. Back to previous page

Intelligent pattern recognition and understanding by integrating probabilistic and symbolic reasoning

Research Project

Project/Area Number 02452281
Research Category

Grant-in-Aid for General Scientific Research (B)

Allocation TypeSingle-year Grants
Research Field Informatics
Research InstitutionKYOTO UNIVERSITY

Principal Investigator

DOSHITA Shuji  Kyoto Univ. Faculty of Engineering Professor, 工学部, 教授 (00025925)

Co-Investigator(Kenkyū-buntansha) ISHIBASHI Hayato  Kyoto Univ. Data Processing Center Assistant Professor, 大型計算機センター, 助手 (70212925)
KAWAHARA Tatsuya  Kyoto Univ. Faculty of Engineering Assistant Professor, 工学部, 助手 (00234104)
KITAZAWA Shigeyoshi  Shizuoka Univ. Faculty of Engineering Associate Professor, 工学部, 助教授 (00109018)
YAMADA Atsushi  Kyoto Univ. Faculty of Engineering Assistant Professor, 工学部, 助手 (20240004)
NISHIDA Toyoaki  Kyoto Univ. Faculty of Engineering Associate Professor, 工学部, 助教授 (70135531)
Project Period (FY) 1990 – 1992
Project Status Completed (Fiscal Year 1992)
Budget Amount *help
¥6,400,000 (Direct Cost: ¥6,400,000)
Fiscal Year 1992: ¥1,500,000 (Direct Cost: ¥1,500,000)
Fiscal Year 1991: ¥1,800,000 (Direct Cost: ¥1,800,000)
Fiscal Year 1990: ¥3,100,000 (Direct Cost: ¥3,100,000)
KeywordsPattern Understanding / Speech Recognition / Speech Understanding / HMM / Context-Free Grammar / Keyword Spotting / Semantic Network / A^*Search / 音声対話 / キ-ワ-ド抽出 / 確率文脈自由文法 / 意味ネットワ-ク / 確率的推論 / 論理的推論 / 自然言語理解 / ベイズ識別器 / ATMS / 概念ネットワ-ク
Research Abstract

For intelligent speech recognition and understanding, we have examined reasoning strategies on several knowledge-levels, and integrated them into speech understanding systems as follows:
(1) Phoneme recognition
We have firstly improved phoneme recognition, which is the base of the whole system. Phoneme HMM based on pair-wise Bayes classifiers is proposed with 27 phoneme recognition rate of 83.1% and 653 word recognition rate of 84.8%.
(2) Syntactic analysis
Syntactic analyzer is developed by integrating probabilistic reasoning and symbolic reasoning on vocabulary and syntax level. Here heuristic search is performed based on prediction by syntax rules and probabilities of HMM. A^*-admissible context-free parsing with word-pair constraints as heuristics is presented.
(3) Keyword spotting
It is possible to make sense of sentences with multiple keywords, without syntax rules. However, conventional method extracts keywords using only the scores of their own, thus insufficient. A new spotting algorithm is presented with assumes logical constraint that the input is a phoneme or word sequence containing target keywords.
(4) Semantic analysis
Network-based semantic analyzer is developed which accepts both N-best word sequences and a keyword lattice and obtains a semantic representation. Here semantic, pragmatic and dialog-level knowledge is integrated and plausible hypothesis is obtained by combining probabilities of candidate words.
(5) Speech understanding system
Two reasoning strategies are implemented on speech understanding systems. One is syntactic-driven which integrates (1), (2) and (4). The other is semantic-driven which integrates (1), (3) and (4). We have evaluated both systems on a task whose vocabulary size is 244 and word perplexity is 80. For grammatical utterances, syntactic-driven approach got an accuracy of 65.5%, while semantic-driven achieved just 44.0%. However, semantic-driven approach is effective for out-of-grammar utterances.

Report

(4 results)
  • 1992 Annual Research Report   Final Research Report Summary
  • 1991 Annual Research Report
  • 1990 Annual Research Report
  • Research Products

    (31 results)

All Other

All Publications (31 results)

  • [Publications] 河原 達也,堂下 修司,北澤 茂良: "判別分析とHMMの統合による不特定話者子音認識" 電子情報通信学会論文誌. J73-D2. 1363-1372 (1990)

    • Description
      「研究成果報告書概要(和文)」より
    • Related Report
      1992 Final Research Report Summary
  • [Publications] T.Kawahara and S.Doshita.: "Phoneme recognition by combining discriminant analysis and HMM" In Proc.of IEEE-ICASSP. 1. 557-560 (1991)

    • Description
      「研究成果報告書概要(和文)」より
    • Related Report
      1992 Final Research Report Summary
  • [Publications] T.KAwahara and S.Doshita.: "HMM based on pair-wise Bayes classifiers" In Proc.of IEEE-ICASSP. 1. 365-368 (1992)

    • Description
      「研究成果報告書概要(和文)」より
    • Related Report
      1992 Final Research Report Summary
  • [Publications] T.Kawahara,S.Matsumoto and S.Doshita.: "A^*-admissible context-free parsing on HMM trellis for speech understanding" In Proc.of Pacific Rim Int'l Conf.on Artificial Intelligence. 2. 1203-1208 (1992)

    • Description
      「研究成果報告書概要(和文)」より
    • Related Report
      1992 Final Research Report Summary
  • [Publications] M.Araki,T.Kawahara,T.Nishida,and S.Doshita.: "Keyword-driven speech parser using dialog-level knowledge" In Proc.of Pacific Rim Int'l Conf.on Artificial Intelligence. 2. 1025-1029 (1992)

    • Description
      「研究成果報告書概要(和文)」より
    • Related Report
      1992 Final Research Report Summary
  • [Publications] 河原 達也,堂下 修司: "対判別に基づく連続型HMMによる音声認識" 電子情報通信学会論文誌. J75-D2. 1641-1648 (1992)

    • Description
      「研究成果報告書概要(和文)」より
    • Related Report
      1992 Final Research Report Summary
  • [Publications] T.Kawahara, T.Ogawa, S.Kitazawa, and S.Doshita.: "Phoneme recognition by combining Bayesian linear discriminations of selected pairs of classes." In Proc. of Int'l Conf. on Spoken Language Processing. 7.8. (1990)

    • Description
      「研究成果報告書概要(欧文)」より
    • Related Report
      1992 Final Research Report Summary
  • [Publications] T.Kawahara and S.Doshita: "Phoneme recognition by combining discriminant analysis and HMM." In Proc. of IEEE-ICASSP. 557-560 (1991)

    • Description
      「研究成果報告書概要(欧文)」より
    • Related Report
      1992 Final Research Report Summary
  • [Publications] P.Fung, T.Kawahara, and S.Doshita: "Unsupervised speaker normalization by speaker Markov model converter for speaker-independent speech recognition." In Proc. of European Conf. on Speech Communication and Technology. 1111-1115 (1991)

    • Description
      「研究成果報告書概要(欧文)」より
    • Related Report
      1992 Final Research Report Summary
  • [Publications] T.Kawahara and S.Doshita: "HMM based on pair-wise Bayes classifiers." In Proc. of IEEE-ICASSP. volume 1. 365-368 (1992)

    • Description
      「研究成果報告書概要(欧文)」より
    • Related Report
      1992 Final Research Report Summary
  • [Publications] T.Kawahara, S.Matsumoto, and S.Doshita: "A^*-admissible context-free parsing on HMM trellis for speech understanding." In Proc. of Pacific Rim Int'l Conf. on Artificial Intelligence. volume 2. 1203-1208 (1992)

    • Description
      「研究成果報告書概要(欧文)」より
    • Related Report
      1992 Final Research Report Summary
  • [Publications] M.Araki, T.Kawahara, T.Nishida, and S.Doshita: "Keyword-driven speech parser using dialog-level knowledge." In Proc. of Pacific Rim Int'l Conf. on Artificial Intelligence. volume 2. 1025-1029 (1992)

    • Description
      「研究成果報告書概要(欧文)」より
    • Related Report
      1992 Final Research Report Summary
  • [Publications] T.Kawahara and S.Doshita: "Comparison of discrete and continuous classifier-based HMM." Journal of Acoustical Society of Japan (E). Vol.13, No.6. 361-367 (1992)

    • Description
      「研究成果報告書概要(欧文)」より
    • Related Report
      1992 Final Research Report Summary
  • [Publications] T.Kawahara and S.Doshita.: "HMM based on pair-wise Bayes classifiers." In Proc.of IEEE-ICASSP. 1. 365-368 (1992)

    • Related Report
      1992 Annual Research Report
  • [Publications] T.Kawahara,S.Matsumoto,and S.Doshita.: "A^*-admissible context-free parsing on HMM trellis for speech understanding." In Proc.of Pacific Rim Int′l Conf.on Artificial Intelligence. 2. 1203-1208 (1992)

    • Related Report
      1992 Annual Research Report
  • [Publications] M.Araki,T.Kawahara,T.Nishida,and S.Doshita.: "Keyword-driven speech parser using dialog-level knowledge." In Proc.of Pacific Rim Int′l Conf.on Artificial Intelligence. 2. 1025-1029 (1992)

    • Related Report
      1992 Annual Research Report
  • [Publications] 河原 達也,堂下 修司.: "対判別に基づく連続型HMMによる音声認識." 電子情報通信学会論文誌. J75DII. 1641-1648 (1992)

    • Related Report
      1992 Annual Research Report
  • [Publications] T.Kawahara and S.Doshita.: "Comparison of discrete and continuous classifier-based HMM." Journal of Acoustical Society of Japan (E). 13. 361-367 (1992)

    • Related Report
      1992 Annual Research Report
  • [Publications] 宗続 敏彦,河原 達也,荒木 雅弘,堂下 修司.: "自由発話理解のためのキーワードスポッティング法." 電子情報通信学会技術報告. SP92-116 (1993)

    • Related Report
      1992 Annual Research Report
  • [Publications] T.Kawahara and S.Doshita: "Phoneme Recognition by Combining Discriminant Analysis and HMM" Proc.of IEEE Int'l Conf.on Acoustics,Speech,& signal Processing. 1. 557-560 (1991)

    • Related Report
      1991 Annual Research Report
  • [Publications] P.Fung,T.Kawahara and S.Doshita: "Unsupervised Speaker Normalization by Speaker Markov Model Converter for SpeakerーIndependent Speech Recognition" Proc.of European Conf.on Speech Communication and Technology. 1111-1115 (1991)

    • Related Report
      1991 Annual Research Report
  • [Publications] T.kawahara and S.Doshita: "HMM based on PairーWise Bayes Classifiers" Proc.of IEEE Int'l Conf.on Acoustics,Speech,& Signal Processing. (1992)

    • Related Report
      1991 Annual Research Report
  • [Publications] 松本 真治,河原 達也,堂下 修司: "語彙・構文・意味制約を統合したA^*探索による会話音声認識" 電子情報通信学会技術報告. SP91ー93. 17-24 (1991)

    • Related Report
      1991 Annual Research Report
  • [Publications] 荒木 雅弘,河原 達也,西田 豊明,堂下 修司: "キ-ワ-ド抽出に基づく意味解析による音声対話システム" 電子情報通信学会技術報告. SP91ー94. 25-32 (1991)

    • Related Report
      1991 Annual Research Report
  • [Publications] 河原 達也,荒木 雅弘,堂下 修司: "leftーto right A^*探索とkeywordーdriven解釈の比較" 電子情報通信学会連続音声認識シンポジウム予稿集. 29-32 (1992)

    • Related Report
      1991 Annual Research Report
  • [Publications] 劉 学敏,西田 豊明,堂下 修司: "統合パ-サによる統合的自然言語解析" 情報処理学会論文誌. 31. 1293-1301 (1990)

    • Related Report
      1990 Annual Research Report
  • [Publications] 河原 達也,堂下 修司,北澤 茂良: "判別分析とHMMの統合による不特定話者子音認識" 電子情報通信学会論文誌. J73ーD2. 1363-1372 (1990)

    • Related Report
      1990 Annual Research Report
  • [Publications] X.Liu,T.Nishida,and S.Doshita: "A Natural Language Understanding System Based on the Integrated Parsing Engine IPE" Proc.of PRICAI'90. 268-273 (1990)

    • Related Report
      1990 Annual Research Report
  • [Publications] T.Kawahara,T.Ogawa,S.Kitazawa,and S.Doshita: "Phoneme Recognition by Combining Bayesian Linear Discriminations of Selected Pairs of Classes" Proc.of International Conference on Spoken Language Processing. 7.8 (1990)

    • Related Report
      1990 Annual Research Report
  • [Publications] 劉 学敏,西田 豊明,堂下 修司: "統合パ-サによるノイズを含んだ文の理解" 情報処理学会自然言語処理研究会報告. 79.2 (1990)

    • Related Report
      1990 Annual Research Report
  • [Publications] 荒木 雅弘,斎藤 隆,佐藤 研治,西田 豊明,堂下 修司: "対話の構造と単語の概念を利用した発話の理解" 第42回情報処理学会全国大会論文集. 3. 61-62 (1991)

    • Related Report
      1990 Annual Research Report

URL: 

Published: 1990-04-01   Modified: 2016-04-21  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi