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

2016 Fiscal Year Annual Research Report

Integration of Event Related Brain Potentials into Speech Recognition Framework

Research Project

Project/Area Number 26870371
Research InstitutionNara Institute of Science and Technology

Principal Investigator

サクリアニ サクティ  奈良先端科学技術大学院大学, 情報科学研究科, 助教 (00395005)

Project Period (FY) 2014-04-01 – 2017-03-31
Keywords音声認識 / 事象関連電位 / 脳波
Outline of Annual Research Achievements

本研究では、人間の知覚に基づき、脳活動計測の一種である事象関連電位(ERP)の研究をASR フレームワークに取り入れ、言語理解における人間の認知プロセスを分析し、自動音声認識(ASR) フレームワークで統合する可能性を検証しました:
(1)ERPによる日本語文での知識と意味のミスマッチ分析
(2)統計的ASRへの知識の統合
最初の年度は、(1)を中心に、研究をしました。特に,本研究では不全要因を(a) システム出力が誤りを含んでいた場合と(b) システム出力は正しいが,コミュニケーションが成立しない場合の2つの場面に分けて考えました.そのうち,前者の例としてASR誤りを,後者の例として未知語の出現を対象とし,それぞれの場面のEEGデータを分析し,コミュニケーション不全要因の検出を試みました.
今年はパート(1)を続け、(2)の研究にも焦点を当てました。近年の深層学習の進展と急速な進歩により、知識のASRへの統合のための深い学習を活用する様々なアプローチも検討しました。また、インドネシアとフランスの他のいくつかの研究機関と共同で共同ワークショップを開催するだけでなく、共同研究を行いました。

  • Research Products

    (26 results)

All 2017 2016 Other

All Int'l Joint Research (3 results) Journal Article (14 results) (of which Int'l Joint Research: 8 results,  Peer Reviewed: 8 results,  Acknowledgement Compliant: 14 results) Presentation (8 results) (of which Int'l Joint Research: 8 results) Funded Workshop (1 results)

  • [Int'l Joint Research] University of Indonesia (UI)/Bandung Institute of Technology (ITB)(Indonesia)

    • Country Name
      Indonesia
    • Counterpart Institution
      University of Indonesia (UI)/Bandung Institute of Technology (ITB)
  • [Int'l Joint Research] International Research Institute MICA(ベトナム)

    • Country Name
      VIET NAM
    • Counterpart Institution
      International Research Institute MICA
  • [Int'l Joint Research] Lab d'Informatique de Grenoble (LIG)/Laboratoire Informatique d'Avignon (LIA)(France)

    • Country Name
      France
    • Counterpart Institution
      Lab d'Informatique de Grenoble (LIG)/Laboratoire Informatique d'Avignon (LIA)
  • [Journal Article] Compressing Recurrent Neural Network with Tensor Train2017

    • Author(s)
      Andros Tjandra, Sakriani Sakti, Satoshi Nakamura
    • Journal Title

      Proceedings of the 2017 International Joint Conference on Neural Networks (IJCNN 2017)

      Volume: 印刷中 Pages: 印刷中

    • Peer Reviewed / Int'l Joint Research / Acknowledgement Compliant
  • [Journal Article] Gated Recurrent Neural Tensor Network2016

    • Author(s)
      Andros Tjandra, Sakriani Sakti, Ruli Manurung, Mirna Adriani, Satoshi Nakamura
    • Journal Title

      Proceedings of The 2016 International Joint Conference on Neural Networks (IJCNN 2016)

      Volume: Vol. 1 Pages: 448-456

    • DOI

      10.1109/IJCNN.2016.7727233

    • Peer Reviewed / Int'l Joint Research / Acknowledgement Compliant
  • [Journal Article] Removing Noise from Event-Related Potentials using a Probabilistic Generative Model with Grouped Covariance Matrices2016

    • Author(s)
      Hayato Maki, Tomoki Toda, Sakriani Sakti, Graham Neubig, Satoshi Nakamura
    • Journal Title

      Proceedings of International Conference of the IEEE Engineering in Medicine and Biology Society

      Volume: Vol. 1 Pages: 3728-3731

    • DOI

      10.1109/EMBC.2016.7591538

    • Peer Reviewed / Int'l Joint Research / Acknowledgement Compliant
  • [Journal Article] Unsupervised Joint Estimation of Grapheme-to-Phoneme Conversion Systems and Acoustic Model Adaptation for Non-Native Speech Recognition2016

    • Author(s)
      Satoshi Tsujioka, Sakriani Sakti, Koichiro Yoshino, Graham Neubig, Satoshi Nakamura
    • Journal Title

      Proceedings of Interspeech 2016

      Volume: Vol. 1 Pages: 3091 - 3095

    • Peer Reviewed / Int'l Joint Research / Acknowledgement Compliant
  • [Journal Article] Supervised Learning of Acoustic Models in a Zero Resource Setting to Improve DPGMM Clustering2016

    • Author(s)
      Michael Heck, Sakriani Sakti, Satoshi Nakamura
    • Journal Title

      Proceedings of Interspeech 2016

      Volume: Vol. 1 Pages: 1310 - 1314

    • Peer Reviewed / Int'l Joint Research / Acknowledgement Compliant
  • [Journal Article] Transferring Emphasis in Speech Translation Using Hard-Attentional Neural Network Models2016

    • Author(s)
      Quoc Truong Do, Sakriani Sakti, Graham Neubig, Satoshi Nakamura
    • Journal Title

      Proceedings of Interspeech 2016

      Volume: Vol. 1 Pages: 2533 - 2537

    • Peer Reviewed / Int'l Joint Research / Acknowledgement Compliant
  • [Journal Article] Iterative Training of A DPGMM-HMM Acoustic Unit Recognizer in A Zero Resource Scenario2016

    • Author(s)
      Michael Heck, Sakriani Sakti, Satoshi Nakamura
    • Journal Title

      Proceedings of IEEE Spoken Language Technology Workshop (SLT 2016)

      Volume: Vol. 1 Pages: 57-63

    • DOI

      10.1109/SLT.2016.7846245

    • Peer Reviewed / Int'l Joint Research / Acknowledgement Compliant
  • [Journal Article] Deep Bottleneck Features And Sound-Dependent i-Vectors for Simultaneous Recognition of Speech and Environmental Sounds2016

    • Author(s)
      Sakriani Sakti, Seiji Kawanishi, Graham Neubig, Koichiro Yoshino, Satoshi Nakamura
    • Journal Title

      Proceedings of IEEE Spoken Language Technology Workshop (SLT 2016)

      Volume: Vol. 1 Pages: 35-42

    • DOI

      10.1109/SLT.2016.7846242

    • Peer Reviewed / Int'l Joint Research / Acknowledgement Compliant
  • [Journal Article] Incongruity Detection on ASR Outputs based on EEG Signals2016

    • Author(s)
      Sakriani Sakti, Yu Odagaki, Takafumi Sasakura, Graham Neubig, Tomoki Toda, Satoshi Nakamura
    • Journal Title

      Proceedings of ASJ 2016

      Volume: Vol. 1 Pages: 83-84

    • Acknowledgement Compliant
  • [Journal Article] A Noise Reduction Method Using Spatial Prior of Event-Related Potentials2016

    • Author(s)
      Hayato Maki, Tomoki Toda, Sakriani Sakti, Graham Neubig, Satoshi Nakamura
    • Journal Title

      Proceedings of ASJ 2016

      Volume: Vol. 1 Pages: 627-628

    • Acknowledgement Compliant
  • [Journal Article] The NAIST ASR for IWSLT: A Multi-architecture DNN System Combination Approach2016

    • Author(s)
      Michael Heck, Quoc Truong Do, Sakriani Sakti, Graham Neubig, Satoshi Nakamura
    • Journal Title

      Proceedings of ASJ 2016

      Volume: Vol. 1 Pages: 337-338

    • Acknowledgement Compliant
  • [Journal Article] Multi-Task Deep Neural Networks for Speech and Environmental Sound Recognition2016

    • Author(s)
      Seiji Kawanishi, Sakriani Sakti, Koichiro Yoshino, Graham Neubig, Satoshi Nakamura
    • Journal Title

      Proceedings of ASJ 2016

      Volume: Vol. 1 Pages: 163-164

    • Acknowledgement Compliant
  • [Journal Article] Non-native Automatic Speech Recognition Utilizing Acoustic Data-driven Pronunciation Learning and Acoustic Model Adaptation2016

    • Author(s)
      Satoshi Tsujioka, Sakriani Sakti, Koichiro Yoshino, Graham Neubig, Satoshi Nakamura
    • Journal Title

      Proceedings of ASJ 2016

      Volume: Vol. 1 Pages: 75-76

    • Acknowledgement Compliant
  • [Journal Article] Exploring Bottleneck Features for Emotional Speech Recognition2016

    • Author(s)
      Kohei Mukaihara, Sakriani Sakti, Koichiro Yoshino, Graham Neubig, Satoshi Nakamura
    • Journal Title

      Proceedings of ASJ 2016

      Volume: Vol. 1 Pages: 161-162

    • Acknowledgement Compliant
  • [Presentation] Compressing Recurrent Neural Network with Tensor Train2017

    • Author(s)
      Andros Tjandra
    • Organizer
      The 2017 International Joint Conference on Neural Networks (IJCNN 2017)
    • Place of Presentation
      Anchorage, Alaska, USA
    • Year and Date
      2017-05-14 – 2017-05-19
    • Int'l Joint Research
  • [Presentation] Iterative Training of A DPGMM-HMM Acoustic Unit Recognizer in A Zero Resource Scenario2016

    • Author(s)
      Michael Heck
    • Organizer
      IEEE Spoken Language Technology Workshop (SLT 2016)
    • Place of Presentation
      San Diego, USA
    • Year and Date
      2016-12-13 – 2016-12-16
    • Int'l Joint Research
  • [Presentation] Deep Bottleneck Features And Sound-Dependent i-Vectors for Simultaneous Recognition of Speech and Environmental Sounds2016

    • Author(s)
      Sakriani Sakti
    • Organizer
      IEEE Spoken Language Technology Workshop (SLT 2016)
    • Place of Presentation
      San Diego, USA
    • Year and Date
      2016-12-13 – 2016-12-16
    • Int'l Joint Research
  • [Presentation] Unsupervised Joint Estimation of Grapheme-to-Phoneme Conversion Systems and Acoustic Model Adaptation for Non-Native Speech Recognition2016

    • Author(s)
      Satoshi Tsujioka
    • Organizer
      Interspeech
    • Place of Presentation
      San Fransisco, USA
    • Year and Date
      2016-09-08 – 2016-09-12
    • Int'l Joint Research
  • [Presentation] Supervised Learning of Acoustic Models in a Zero Resource Setting to Improve DPGMM Clustering2016

    • Author(s)
      Michael Heck
    • Organizer
      Interspeech
    • Place of Presentation
      San Fransisco, USA
    • Year and Date
      2016-09-08 – 2016-09-12
    • Int'l Joint Research
  • [Presentation] Transferring Emphasis in Speech Translation Using Hard-Attentional Neural Network Models2016

    • Author(s)
      Quoc Truong Do
    • Organizer
      Interspeech
    • Place of Presentation
      San Fransisco, USA
    • Year and Date
      2016-09-08 – 2016-09-12
    • Int'l Joint Research
  • [Presentation] Removing Noise from Event-Related Potentials using a Probabilistic Generative Model with Grouped Covariance Matrices2016

    • Author(s)
      Hayato Maki
    • Organizer
      International Conference of the IEEE Engineering in Medicine and Biology Society
    • Place of Presentation
      Orlando, Florida, USA
    • Year and Date
      2016-08-16 – 2016-08-20
    • Int'l Joint Research
  • [Presentation] Gated Recurrent Neural Tensor Network2016

    • Author(s)
      Sakriani Sakti
    • Organizer
      The 2016 International Joint Conference on Neural Networks (IJCNN 2016)
    • Place of Presentation
      Vancouver, Canada
    • Year and Date
      2016-07-24 – 2016-07-29
    • Int'l Joint Research
  • [Funded Workshop] The 5th International Workshop on Spoken Language Technologies for Under-resourced Languages (SLTU'16)2016

    • Place of Presentation
      Yogyakarta, Indonesia
    • Year and Date
      2016-05-09 – 2016-05-12

URL: 

Published: 2018-01-16   Modified: 2022-02-16  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi