Path: Top > Gray Literatures > S > Students > 2005

SPEECH SIGNAL ANALYSIS USING WAVELET TRANSFORM MODULUS MAXIMA

Proceeding Konferensi Nasional Teknologi Informasi & Komunikasi untuk Indonesia, Institut Teknologi Bandung, 3-4 Mei 2005
Gray literature from JBPTITBPP / 2007-02-14 18:47:43
Oleh : Hendryk B. Nugraha, DSP Research and Technology Group (hendryk@ieee.org)
Dibuat : 2005-05-03, dengan 1 file

Keyword :

Speech, Turbulence, Fractal, WTMM


Abstrak:

Linear acoustics and time-invariant system has been used for decades to model the speech signals, ignoring the true nonlinear physics of those signals. Recent research suggest that speech production may be a nonlinear process which motivate interest in the area of nonlinear analysis of speech signals.

The dynamics of speech airflow might create small or large degrees of turbulence during production of speech sounds by the human vocal-tract system. This multiscale structure of turbulence can in some cases be analyze by fractal model., Wavelets provide new techniques which deserve a special attention in the context of fractal analysis and synthesis because they can be used to extract microscopic information about their scaling properties. Wavelet Transform Modulus Maxima (WTMM) allows to locally analyse the singular behaviour of signals, in particular to trace the singularities and to characterize them quantitatively.

Motivated turbulence model in speech signal generation, fractal analysis of turbulence and WTMM approach for fractal analysis, we proposed an algorithm to analyze speech signal using WTMM.

Deskripsi Alternatif :

Abstrak:

Linear acoustics and time-invariant system has been used for decades to model the speech signals, ignoring the true nonlinear physics of those signals. Recent research suggest that speech production may be a nonlinear process which motivate interest in the area of nonlinear analysis of speech signals.

The dynamics of speech airflow might create small or large degrees of turbulence during production of speech sounds by the human vocal-tract system. This multiscale structure of turbulence can in some cases be analyze by fractal model., Wavelets provide new techniques which deserve a special attention in the context of fractal analysis and synthesis because they can be used to extract microscopic information about their scaling properties. Wavelet Transform Modulus Maxima (WTMM) allows to locally analyse the singular behaviour of signals, in particular to trace the singularities and to characterize them quantitatively.

Motivated turbulence model in speech signal generation, fractal analysis of turbulence and WTMM approach for fractal analysis, we proposed an algorithm to analyze speech signal using WTMM.


Beri Komentar ?#(0) | Bookmark

PropertiNilai Properti
ID PublisherJBPTITBPP
OrganisasiD
Nama KontakUPT Perpustakaan ITB
AlamatJl. Ganesha 10
KotaBandung
DaerahJawa Barat
NegaraIndonesia
Telepon62-22-2509118, 2500089
Fax62-22-2500089
E-mail Administratordigilib@lib.itb.ac.id
E-mail CKOinfo@lib.itb.ac.id

Print ...

Kontributor...

  • Penulis 2
    Armein Z. R. Langi, Dr.Ir., M.Sc.
    langi@lss.ee.itb.ac.id
    langi@paume.itb.ac.id
    langi@ieee.org

    Scan:
    Ena Sukmana, S.Sos. (2006-02-20), Editor:

File PDF...