• DocumentCode
    1712682
  • Title

    Analysis of glottal signals for speaker information

  • Author

    K, Ramesh ; Pradhan, Gayadhar ; Prasanna, S.R.Mahadeva

  • Author_Institution
    Dept. of EEE, IIT Guwahati, 781039, India
  • fYear
    2013
  • Firstpage
    1
  • Lastpage
    4
  • Abstract
    This work analyzes the glottal signals recorded from different speakers for speaker information and describes some speaker specific cues. The glottal activity (GA) regions are detected by exploiting the strength of excitation and periodic nature of glottal signal. The identified GA regions are further processed to segment into nearly periodic smaller segments by detecting peaks in individual glottal cycles and using them as anchor points. The periodic segments are further divided into four smallest segments by considering the peaks and zero crossings. The speaker information present at the GA region, periodic segment and the four smallest segments are analyzed and illustrated. From the experimental studies the speaker information manifested at the four smallest segments show less intra speaker variability and more inter speaker variability. The four smallest segments of the periodic segments are therefore proposed as the speaker specific cues for the development of speaker recognition systems.
  • Keywords
    Data mining; Feature extraction; Mel frequency cepstral coefficient; Production; Speaker recognition; Speech; Testing; Glottal signal; four smallest segments; glottal activity region; periodic segment; speaker information;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Communications (NCC), 2013 National Conference on
  • Conference_Location
    New Delhi, India
  • Print_ISBN
    978-1-4673-5950-4
  • Electronic_ISBN
    978-1-4673-5951-1
  • Type

    conf

  • DOI
    10.1109/NCC.2013.6487996
  • Filename
    6487996