EFFECTIVE PARAMETER FOR RECOGNIZING VOWEL IN CHINESE SPEECH - VOCAL TRACT LENGTH.

被引:0
|
作者
Chai, Peiqi
机构
关键词
SIGNAL PROCESSING - Measurements;
D O I
暂无
中图分类号
学科分类号
摘要
This paper is based on the similarity of the anatomical structures of the vocal organs among adults and the regularity of vocal tract length. The purpose is to distinguish between different vowels in Chinese speech. We have discovered that vocal tract lengths of spoken vowels a, o, u, i, e have fixed length rates and their absolute lengths are quite certain. Therefore this characteristic is useful for automatic identification of speech sounds produced by arbitrary speakers. The vocal tract length is not directly available from the speech signal. A hypothesis which uses the fourth or the fifth formant frequency to determine the vocal tract length is proposed. There is very good agreement between our experimental length parameter and Fant's reference value.
引用
收藏
页码:177 / 181
相关论文
共 33 条
  • [21] Vocal tract length normalization for speaker independent acoustic-to-articulatory speech inversion
    Sivaraman, Ganesh
    Mitra, Vikramjit
    Nam, Hosung
    Tiede, Mark
    Espy-Wilson, Carol
    17TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION (INTERSPEECH 2016), VOLS 1-5: UNDERSTANDING SPEECH PROCESSING IN HUMANS AND MACHINES, 2016, : 455 - 459
  • [22] Vocal tract length normalization using rapid maximum-likelihood estimation for speech recognition
    Emori, Tadashi
    Shinoda, Koichi
    Systems and Computers in Japan, 2002, 33 (05): : 30 - 40
  • [23] A statistical, formant-pattern model for segregating vowel type and vocal-tract length in developmental formant data
    Turner, Richard E.
    Walters, Thomas C.
    Monaghan, Jessica J. M.
    Patterson, Roy D.
    JOURNAL OF THE ACOUSTICAL SOCIETY OF AMERICA, 2009, 125 (04): : 2374 - 2386
  • [24] Vocal Tract Length Normalization and Sub-Band Spectral Subtraction Based Robust Assamese Vowel Recognition System
    Gogoi, Swapnanil
    Bhattacharjee, Utpal
    2017 INTERNATIONAL CONFERENCE ON COMPUTING METHODOLOGIES AND COMMUNICATION (ICCMC), 2017, : 32 - 35
  • [25] Feature compensation based on the normalization of vocal tract length for the improvement of emotion-affected speech recognition
    Masoud Geravanchizadeh
    Elnaz Forouhandeh
    Meysam Bashirpour
    EURASIP Journal on Audio, Speech, and Music Processing, 2021
  • [26] Feature compensation based on the normalization of vocal tract length for the improvement of emotion-affected speech recognition
    Geravanchizadeh, Masoud
    Forouhandeh, Elnaz
    Bashirpour, Meysam
    EURASIP JOURNAL ON AUDIO SPEECH AND MUSIC PROCESSING, 2021, 2021 (01)
  • [27] VOCAL TRACT LENGTH NORMALISATION APPROACHES TO DNN-BASED CHILDREN'S AND ADULTS' SPEECH RECOGNITION
    Serizel, Romain
    Giuliani, Diego
    2014 IEEE WORKSHOP ON SPOKEN LANGUAGE TECHNOLOGY SLT 2014, 2014, : 135 - 140
  • [28] Improved vocal tract length perturbation for a state-of-the-art end-to-end speech recognition system
    Kim, Chanwoo
    Shin, Minkyu
    Garg, Abhinav
    Gowda, Dhananjaya
    INTERSPEECH 2019, 2019, : 739 - 743
  • [29] Research of whispered speech vocal tract system conversion based on universal background model and effective Gaussian components
    CHEN Xueqin
    ZHAO Heming
    ChineseJournalofAcoustics, 2013, 32 (04) : 400 - 410