Blind reverberation energy estimation using exponential averaging with attack and release time constants for hearing aids

Kotoyo Nozaki, Yusuke Ikeda, Yasuhiro Oikawa, Yoh ichi Fujisaka, Masahiro Sunohara

    Research output: Contribution to journalArticle

    Abstract

    Dereverberation signal processing is necessary for hearing-aid systems because reverberation degrades speech intelligibility in reverberant environments. In general, hearing-aid systems require low computational cost and real-time signal processing. The spectral subtraction (SS) method is a simple and frequently used technique which is used not only for noise-reduction but for dereverberation as well. To perform dereverberation methods based on SS for hearing aids, the reverberation energy must be estimated or measured. In this paper, the SS-based blind estimation of reverberation energy in a single-channel speech signal is proposed by using exponential averaging with attack and release time constants. The estimation error, which is the difference between the true and estimated reverberation energy, was used for evaluation. The estimation error of the proposed method was compared with the results of the method proposed by Lebart et al., which is a well-known non-blind SS-based dereverberation method. According to the results, the reverberation energy was more accurately estimated when the reverberation time was longer than 0.6 s, and the estimation error of the proposed method was approximately half of that of the well-known non-blind method by Lebart.

    Original languageEnglish
    Pages (from-to)106-113
    Number of pages8
    JournalApplied Acoustics
    Volume142
    DOIs
    Publication statusPublished - 2018 Dec 15

    Fingerprint

    reverberation
    hearing
    attack
    time constant
    subtraction
    energy
    signal processing
    intelligibility
    time signals
    noise reduction
    costs
    evaluation

    Keywords

    • Blind estimation
    • Exponential averaging
    • Hearing aids
    • Reverberant speech
    • Spectral subtraction

    ASJC Scopus subject areas

    • Acoustics and Ultrasonics

    Cite this

    Blind reverberation energy estimation using exponential averaging with attack and release time constants for hearing aids. / Nozaki, Kotoyo; Ikeda, Yusuke; Oikawa, Yasuhiro; Fujisaka, Yoh ichi; Sunohara, Masahiro.

    In: Applied Acoustics, Vol. 142, 15.12.2018, p. 106-113.

    Research output: Contribution to journalArticle

    Nozaki, Kotoyo ; Ikeda, Yusuke ; Oikawa, Yasuhiro ; Fujisaka, Yoh ichi ; Sunohara, Masahiro. / Blind reverberation energy estimation using exponential averaging with attack and release time constants for hearing aids. In: Applied Acoustics. 2018 ; Vol. 142. pp. 106-113.
    @article{fe89218f9b124e9f82a88631946b28db,
    title = "Blind reverberation energy estimation using exponential averaging with attack and release time constants for hearing aids",
    abstract = "Dereverberation signal processing is necessary for hearing-aid systems because reverberation degrades speech intelligibility in reverberant environments. In general, hearing-aid systems require low computational cost and real-time signal processing. The spectral subtraction (SS) method is a simple and frequently used technique which is used not only for noise-reduction but for dereverberation as well. To perform dereverberation methods based on SS for hearing aids, the reverberation energy must be estimated or measured. In this paper, the SS-based blind estimation of reverberation energy in a single-channel speech signal is proposed by using exponential averaging with attack and release time constants. The estimation error, which is the difference between the true and estimated reverberation energy, was used for evaluation. The estimation error of the proposed method was compared with the results of the method proposed by Lebart et al., which is a well-known non-blind SS-based dereverberation method. According to the results, the reverberation energy was more accurately estimated when the reverberation time was longer than 0.6 s, and the estimation error of the proposed method was approximately half of that of the well-known non-blind method by Lebart.",
    keywords = "Blind estimation, Exponential averaging, Hearing aids, Reverberant speech, Spectral subtraction",
    author = "Kotoyo Nozaki and Yusuke Ikeda and Yasuhiro Oikawa and Fujisaka, {Yoh ichi} and Masahiro Sunohara",
    year = "2018",
    month = "12",
    day = "15",
    doi = "10.1016/j.apacoust.2018.08.010",
    language = "English",
    volume = "142",
    pages = "106--113",
    journal = "Applied Acoustics",
    issn = "0003-682X",
    publisher = "Elsevier Limited",

    }

    TY - JOUR

    T1 - Blind reverberation energy estimation using exponential averaging with attack and release time constants for hearing aids

    AU - Nozaki, Kotoyo

    AU - Ikeda, Yusuke

    AU - Oikawa, Yasuhiro

    AU - Fujisaka, Yoh ichi

    AU - Sunohara, Masahiro

    PY - 2018/12/15

    Y1 - 2018/12/15

    N2 - Dereverberation signal processing is necessary for hearing-aid systems because reverberation degrades speech intelligibility in reverberant environments. In general, hearing-aid systems require low computational cost and real-time signal processing. The spectral subtraction (SS) method is a simple and frequently used technique which is used not only for noise-reduction but for dereverberation as well. To perform dereverberation methods based on SS for hearing aids, the reverberation energy must be estimated or measured. In this paper, the SS-based blind estimation of reverberation energy in a single-channel speech signal is proposed by using exponential averaging with attack and release time constants. The estimation error, which is the difference between the true and estimated reverberation energy, was used for evaluation. The estimation error of the proposed method was compared with the results of the method proposed by Lebart et al., which is a well-known non-blind SS-based dereverberation method. According to the results, the reverberation energy was more accurately estimated when the reverberation time was longer than 0.6 s, and the estimation error of the proposed method was approximately half of that of the well-known non-blind method by Lebart.

    AB - Dereverberation signal processing is necessary for hearing-aid systems because reverberation degrades speech intelligibility in reverberant environments. In general, hearing-aid systems require low computational cost and real-time signal processing. The spectral subtraction (SS) method is a simple and frequently used technique which is used not only for noise-reduction but for dereverberation as well. To perform dereverberation methods based on SS for hearing aids, the reverberation energy must be estimated or measured. In this paper, the SS-based blind estimation of reverberation energy in a single-channel speech signal is proposed by using exponential averaging with attack and release time constants. The estimation error, which is the difference between the true and estimated reverberation energy, was used for evaluation. The estimation error of the proposed method was compared with the results of the method proposed by Lebart et al., which is a well-known non-blind SS-based dereverberation method. According to the results, the reverberation energy was more accurately estimated when the reverberation time was longer than 0.6 s, and the estimation error of the proposed method was approximately half of that of the well-known non-blind method by Lebart.

    KW - Blind estimation

    KW - Exponential averaging

    KW - Hearing aids

    KW - Reverberant speech

    KW - Spectral subtraction

    UR - http://www.scopus.com/inward/record.url?scp=85052458734&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=85052458734&partnerID=8YFLogxK

    U2 - 10.1016/j.apacoust.2018.08.010

    DO - 10.1016/j.apacoust.2018.08.010

    M3 - Article

    AN - SCOPUS:85052458734

    VL - 142

    SP - 106

    EP - 113

    JO - Applied Acoustics

    JF - Applied Acoustics

    SN - 0003-682X

    ER -