話者認識技術を用いた性同一性症者(MtF)の音声に対する男声度・女声度の自動推定とその臨床応用  [in Japanese] Automatic estimation of femininity of MtF's speech using speaker recognition technique and its clinical application  [in Japanese]

Search this Article

Author(s)

Abstract

著者らは男性から女性へ性別の移行を希望する性同一性障害者(Male-to-Female transgendered/transsexual=MtF)に対して、声を女性化させるためのtranssexual voice therapyを行っており、今回の発表では話者認識技術を用いて知覚的女性度を推定するシステムの臨床応用について検討する。このシステムは、声道特性と音源特性それぞれについて、男声モデル・女声モデルを持ち、各特性別に入力音声の女声度を推定、聴取実験により女性と判定される率(知覚的女声度)の予測値を算出する。上記のシステムを実際の臨床で用いた結果、声道形状を変えながらピッチをあげて女声をつくる方略の完成度を知ることはできるが、発話スタイルの動的制御に基づく女声の生成方略には対応できておらず、今後の検討の課題であることがわかった。

This study reports the use of a technique of automatic estimation of perceptual femininity in voice therapy for Male-to-Female transgender/transsexuals(MtF). The technique is based on speaker recognition techniques. Male models and female models are separately trained for vocal tract shape and pitch distribution. Using the four models, i.e., two genders and two parameters, perceptual femininity that is obtained through listening tests, is automatically estimated. The femininity estimation system is applied in actual voice therapy of MtF clients. Although it was accepted well to the clients, the estimation results sometimes are not fit to the femininity perceived by the therapist. This is mainly because the system captures only static acoustic properties of speech. This implies that the therapist can focus more on dynamic control of speech production in the therapy.

Journal

  • IEICE technical report

    IEICE technical report 105(686), 29-34, 2006-03-28

    The Institute of Electronics, Information and Communication Engineers

References:  9

Cited by:  1

Codes

  • NII Article ID (NAID)
    110004680238
  • NII NACSIS-CAT ID (NCID)
    AN10013221
  • Text Lang
    JPN
  • Article Type
    Journal Article
  • ISSN
    09135685
  • NDL Article ID
    7880189
  • NDL Source Classification
    ZN33(科学技術--電気工学・電気機械工業--電子工学・電気通信)
  • NDL Call No.
    Z16-940
  • Data Source
    CJP  CJPref  NDL  NII-ELS 
Page Top