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

A Study on Improvements of Noise Rebustness for HMM-based Speech Recognition

Research Project

Project/Area Number 05680294
Research Category

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

Allocation TypeSingle-year Grants
Research Field Intelligent informatics
Research InstitutionShinshu University

Principal Investigator

MATSUMOTO Hiroshi  Shinshu University, Faculty of Engineering, Professor, 工学部・電気電子工学科, 教授 (60005452)

Project Period (FY) 1993 – 1994
Project Status Completed (Fiscal Year 1994)
Budget Amount *help
¥1,800,000 (Direct Cost: ¥1,800,000)
Fiscal Year 1994: ¥300,000 (Direct Cost: ¥300,000)
Fiscal Year 1993: ¥1,500,000 (Direct Cost: ¥1,500,000)
KeywordsSpeech Recognition / Hidden Markov Model / Noisy Environments / Frequency Weighting / Euclidean Distance / Noise Robustness
Research Abstract

In order to realize robust continuous density Hidden Markov Models (HMM) for noisy speech recognition, this study develops a frequency-weighted HMM based on the human auditory characteristics which is seseitive to formant peaks in high SNR frequency region. In this HMM,the covariance matrices of Gaussian probability density functions are fixed to the inverse of frequency weighting matrices in order to utilize the robustness of group delay spectra and also to incorporate their relative perceptual importance in frequency domain into HMM.Several frequency weighting functions and the scaling methods of frequency weighting matrices are examined using the international data base of NOISEX-92. The results of word recognition tests are summarized as follows.
(1) The smoothed power spectrum derived from each mean vector gives the most robust HMM.
(2) The optimum scaling to convert the weighting matrices to the covariance matrices is such that the sum of weighting coefficients is equal to one or the determinants of the converted covariances are 50 to 150 times larger than those of initial HMMs.
(3) A larger number of states is required to attain the robustness in the frequency-weighted HMM.
(4) Adaptive preemphasis improves the robustness to noises which have less energy in the high frequency region.
(5) The frequency-weighted HMM attains SNR gains of 6 to 12 dB over a standard diagonal HMM for white, pink, and car noises.
(6) Even when preprocessing the noisy speech by the standard noise reduction method of spectral subtraction, the frequency weighted HMM attains about 10% higher recognition scores in very low SNR condition than the conventional HMM.

Report

(3 results)
  • 1994 Annual Research Report   Final Research Report Summary
  • 1993 Annual Research Report
  • Research Products

    (17 results)

All Other

All Publications (17 results)

  • [Publications] H.Matsumoto & Y.Nimura: "A Frequency-weighted continuous density HMM for noisy speech recognition" Proc.of Int.Conf.on Spoken Language Processing. 1007-1010 (1994)

    • Description
      「研究成果報告書概要(和文)」より
    • Related Report
      1994 Final Research Report Summary
  • [Publications] H.Masumoto: "Robust speech recognition in noisy environments" Proc.of Int.Workshop on Human Interface Technology. 1-8 (1994)

    • Description
      「研究成果報告書概要(和文)」より
    • Related Report
      1994 Final Research Report Summary
  • [Publications] H.Matsumoto & Y.Nimura: "An improved spectral subtraction with smoothing for noisy speech recognition" Proc.of 15th Int.Congress on Acoustics. 1-4 (1995)

    • Description
      「研究成果報告書概要(和文)」より
    • Related Report
      1994 Final Research Report Summary
  • [Publications] H.Matsumoto & Y.Nimura: "A frequency-weighted continuous density HMM for noisy speech recognition, Proc.of Int. Conf.on Spoken Language Processing" 1007-1010 (1994)

    • Description
      「研究成果報告書概要(欧文)」より
    • Related Report
      1994 Final Research Report Summary
  • [Publications] H.Matsumoto: "Robust speech recognition in noisy environments" Proc.of Int. Workshop on Human Interface Technology. 1-8 (1994)

    • Description
      「研究成果報告書概要(欧文)」より
    • Related Report
      1994 Final Research Report Summary
  • [Publications] H.Matsumoto & Y.Nimura: "An improved spectral subtraction with smoothing for noisy speech recognition" Proc.of 15th Int. Congress on Acoustics. 1-8 (1995)

    • Description
      「研究成果報告書概要(欧文)」より
    • Related Report
      1994 Final Research Report Summary
  • [Publications] H.Matsumoto & Y.Nimura: "A frequency-weighted continuous density HMM for noisy speech recognition" Proc. of Int.Conf.on Spoken Language Processing. 1007-1010 (1994)

    • Related Report
      1994 Annual Research Report
  • [Publications] H.Matsumoto: "Robust speech recognition in noisy environments" Proc. of Int. Workshop on Human Interfase Technology. 1-8 (1994)

    • Related Report
      1994 Annual Research Report
  • [Publications] H.Matsumoto & Y.Nimura: "An improved spectral subtraction with smoothing of noisy speech recognition" Proc. of 15th Int. Congress on Acoustics. 1-4 (1995)

    • Related Report
      1994 Annual Research Report
  • [Publications] 二村善則、松本 弘: "周波数平滑化スペクトルサブトラクションのNOISEX-92による評価" 電子情報通信学会講演論文集. 491-492 (1994)

    • Related Report
      1994 Annual Research Report
  • [Publications] 永山 亮、松本 弘: "音声認識における雑音付加HMMの自動生成" 日本音響学会講演論文集. 59-60 (1995)

    • Related Report
      1994 Annual Research Report
  • [Publications] 妹背博之、松本弘: "雑音下音声認識における周波数重み付けHMMの改良" 日本音響学会講演論文集. 141-142 (1993)

    • Related Report
      1993 Annual Research Report
  • [Publications] 二村善則、松本弘: "各種BPF距離尺度に対するスペクトルサブトラクションの検討" 日本音響学会講演論文集. 149-150 (1993)

    • Related Report
      1993 Annual Research Report
  • [Publications] 永山亮、松本弘: "HMM音声認識におけるデルタケプストラムの雑音への適応" 電子情報通信学会信越支部大会講演論文集. 61-62 (1993)

    • Related Report
      1993 Annual Research Report
  • [Publications] 妹背博之、松本弘: "雑音下音声認識における周波数重み付けHMMの改良と評価" 電子情報通信学会音声研究会資料. SP93-107. 25-32 (1993)

    • Related Report
      1993 Annual Research Report
  • [Publications] 二村善則、松本弘: "スペクトルサブトラクション法の改良と非定常雑音による評価" 日本音響学会講演論文集. 5-6 (1994)

    • Related Report
      1993 Annual Research Report
  • [Publications] 永山亮、松本弘: "スペクトル系列の生成に基づく雑音付加HMMの合成" 日本音響学会講演論文集. 7-8 (1994)

    • Related Report
      1993 Annual Research Report

URL: 

Published: 1993-04-01   Modified: 2016-04-21  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi