CENSREC-4

An evaluation framework for distant-talking speech recognition in reverberant environments

Takahiro Fukumori, Takanobu Nishiura, Masato Nakayama, Yuki Denda, Norihide Kitaoka, Takeshi Yamada, Kazumasa Yamamoto, Satoru Tsuge, Masakiyo Fujimoto, Tetsuya Takiguchi, Chiyomi Miyajima, Satoshi Tamura, Tetsuji Ogawa, Shigeki Matsuda, Shingo Kuroiwa, Kazuya Takeda, Satoshi Nakamura

    Research output: Contribution to journalArticle

    2 Citations (Scopus)

    Abstract

    We have been distributing a new collection of databases and evaluation tools called CENSREC-4, which is a framework for evaluating distant-talking speech in reverberant environments. The data contained in CENSREC-4 are connected digit utterances as in CENSREC-1. Two subsets are included in the data: "basic data sets" and "extra data sets." The basic data sets are used for evaluating the room impulse response-convolved speech data to simulate the various reverberations. The extra data sets consist of simulated data and corresponding real recorded data. Evaluation tools are presently only provided for the basic data sets and will be delivered to the extra data sets in the future. The task of CENSREC-4 with a basic data set appears simple; however, the results of experiments prove that CENSREC-4 provides a challenging reverberation speech-recognition task, in the sense that a traditional technique to improve recognition and a widely used criterion to represent the difficulty of recognition deliver poor performance. Within this context, this common framework can be an important step toward the future evolution of reverberant speech-recognition methodologies.

    Original languageEnglish
    Pages (from-to)201-210
    Number of pages10
    JournalAcoustical Science and Technology
    Volume32
    Issue number5
    DOIs
    Publication statusPublished - 2011

    Fingerprint

    talking
    speech recognition
    reverberation
    distributing
    digits
    evaluation
    set theory
    rooms
    impulses
    methodology

    Keywords

    • Evaluation framework
    • Reverberant speech database
    • Reverberant speech recognition
    • Room impulse response
    • Various recording environments

    ASJC Scopus subject areas

    • Acoustics and Ultrasonics

    Cite this

    Fukumori, T., Nishiura, T., Nakayama, M., Denda, Y., Kitaoka, N., Yamada, T., ... Nakamura, S. (2011). CENSREC-4: An evaluation framework for distant-talking speech recognition in reverberant environments. Acoustical Science and Technology, 32(5), 201-210. https://doi.org/10.1250/ast.32.201

    CENSREC-4 : An evaluation framework for distant-talking speech recognition in reverberant environments. / Fukumori, Takahiro; Nishiura, Takanobu; Nakayama, Masato; Denda, Yuki; Kitaoka, Norihide; Yamada, Takeshi; Yamamoto, Kazumasa; Tsuge, Satoru; Fujimoto, Masakiyo; Takiguchi, Tetsuya; Miyajima, Chiyomi; Tamura, Satoshi; Ogawa, Tetsuji; Matsuda, Shigeki; Kuroiwa, Shingo; Takeda, Kazuya; Nakamura, Satoshi.

    In: Acoustical Science and Technology, Vol. 32, No. 5, 2011, p. 201-210.

    Research output: Contribution to journalArticle

    Fukumori, T, Nishiura, T, Nakayama, M, Denda, Y, Kitaoka, N, Yamada, T, Yamamoto, K, Tsuge, S, Fujimoto, M, Takiguchi, T, Miyajima, C, Tamura, S, Ogawa, T, Matsuda, S, Kuroiwa, S, Takeda, K & Nakamura, S 2011, 'CENSREC-4: An evaluation framework for distant-talking speech recognition in reverberant environments', Acoustical Science and Technology, vol. 32, no. 5, pp. 201-210. https://doi.org/10.1250/ast.32.201
    Fukumori, Takahiro ; Nishiura, Takanobu ; Nakayama, Masato ; Denda, Yuki ; Kitaoka, Norihide ; Yamada, Takeshi ; Yamamoto, Kazumasa ; Tsuge, Satoru ; Fujimoto, Masakiyo ; Takiguchi, Tetsuya ; Miyajima, Chiyomi ; Tamura, Satoshi ; Ogawa, Tetsuji ; Matsuda, Shigeki ; Kuroiwa, Shingo ; Takeda, Kazuya ; Nakamura, Satoshi. / CENSREC-4 : An evaluation framework for distant-talking speech recognition in reverberant environments. In: Acoustical Science and Technology. 2011 ; Vol. 32, No. 5. pp. 201-210.
    @article{e468e5191ddd4ae3a8b91da2972c5b79,
    title = "CENSREC-4: An evaluation framework for distant-talking speech recognition in reverberant environments",
    abstract = "We have been distributing a new collection of databases and evaluation tools called CENSREC-4, which is a framework for evaluating distant-talking speech in reverberant environments. The data contained in CENSREC-4 are connected digit utterances as in CENSREC-1. Two subsets are included in the data: {"}basic data sets{"} and {"}extra data sets.{"} The basic data sets are used for evaluating the room impulse response-convolved speech data to simulate the various reverberations. The extra data sets consist of simulated data and corresponding real recorded data. Evaluation tools are presently only provided for the basic data sets and will be delivered to the extra data sets in the future. The task of CENSREC-4 with a basic data set appears simple; however, the results of experiments prove that CENSREC-4 provides a challenging reverberation speech-recognition task, in the sense that a traditional technique to improve recognition and a widely used criterion to represent the difficulty of recognition deliver poor performance. Within this context, this common framework can be an important step toward the future evolution of reverberant speech-recognition methodologies.",
    keywords = "Evaluation framework, Reverberant speech database, Reverberant speech recognition, Room impulse response, Various recording environments",
    author = "Takahiro Fukumori and Takanobu Nishiura and Masato Nakayama and Yuki Denda and Norihide Kitaoka and Takeshi Yamada and Kazumasa Yamamoto and Satoru Tsuge and Masakiyo Fujimoto and Tetsuya Takiguchi and Chiyomi Miyajima and Satoshi Tamura and Tetsuji Ogawa and Shigeki Matsuda and Shingo Kuroiwa and Kazuya Takeda and Satoshi Nakamura",
    year = "2011",
    doi = "10.1250/ast.32.201",
    language = "English",
    volume = "32",
    pages = "201--210",
    journal = "Acoustical Science and Technology",
    issn = "1346-3969",
    publisher = "Acoustical Society of Japan",
    number = "5",

    }

    TY - JOUR

    T1 - CENSREC-4

    T2 - An evaluation framework for distant-talking speech recognition in reverberant environments

    AU - Fukumori, Takahiro

    AU - Nishiura, Takanobu

    AU - Nakayama, Masato

    AU - Denda, Yuki

    AU - Kitaoka, Norihide

    AU - Yamada, Takeshi

    AU - Yamamoto, Kazumasa

    AU - Tsuge, Satoru

    AU - Fujimoto, Masakiyo

    AU - Takiguchi, Tetsuya

    AU - Miyajima, Chiyomi

    AU - Tamura, Satoshi

    AU - Ogawa, Tetsuji

    AU - Matsuda, Shigeki

    AU - Kuroiwa, Shingo

    AU - Takeda, Kazuya

    AU - Nakamura, Satoshi

    PY - 2011

    Y1 - 2011

    N2 - We have been distributing a new collection of databases and evaluation tools called CENSREC-4, which is a framework for evaluating distant-talking speech in reverberant environments. The data contained in CENSREC-4 are connected digit utterances as in CENSREC-1. Two subsets are included in the data: "basic data sets" and "extra data sets." The basic data sets are used for evaluating the room impulse response-convolved speech data to simulate the various reverberations. The extra data sets consist of simulated data and corresponding real recorded data. Evaluation tools are presently only provided for the basic data sets and will be delivered to the extra data sets in the future. The task of CENSREC-4 with a basic data set appears simple; however, the results of experiments prove that CENSREC-4 provides a challenging reverberation speech-recognition task, in the sense that a traditional technique to improve recognition and a widely used criterion to represent the difficulty of recognition deliver poor performance. Within this context, this common framework can be an important step toward the future evolution of reverberant speech-recognition methodologies.

    AB - We have been distributing a new collection of databases and evaluation tools called CENSREC-4, which is a framework for evaluating distant-talking speech in reverberant environments. The data contained in CENSREC-4 are connected digit utterances as in CENSREC-1. Two subsets are included in the data: "basic data sets" and "extra data sets." The basic data sets are used for evaluating the room impulse response-convolved speech data to simulate the various reverberations. The extra data sets consist of simulated data and corresponding real recorded data. Evaluation tools are presently only provided for the basic data sets and will be delivered to the extra data sets in the future. The task of CENSREC-4 with a basic data set appears simple; however, the results of experiments prove that CENSREC-4 provides a challenging reverberation speech-recognition task, in the sense that a traditional technique to improve recognition and a widely used criterion to represent the difficulty of recognition deliver poor performance. Within this context, this common framework can be an important step toward the future evolution of reverberant speech-recognition methodologies.

    KW - Evaluation framework

    KW - Reverberant speech database

    KW - Reverberant speech recognition

    KW - Room impulse response

    KW - Various recording environments

    UR - http://www.scopus.com/inward/record.url?scp=80052431269&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=80052431269&partnerID=8YFLogxK

    U2 - 10.1250/ast.32.201

    DO - 10.1250/ast.32.201

    M3 - Article

    VL - 32

    SP - 201

    EP - 210

    JO - Acoustical Science and Technology

    JF - Acoustical Science and Technology

    SN - 1346-3969

    IS - 5

    ER -