Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Denoising human speech signals using chaoslike features.

R Hegger1, H Kantz, L Matassini

  • 1Max-Planck-Institut für Physik komplexer Systeme, Nöthnitzer Strasse 38, D 01187 Dresden, Germany.

Physical Review Letters
|October 6, 2000
PubMed
Summary

A novel noise reduction technique for deterministic chaotic signals was successfully adapted for human speech processing. This method leverages speech signal structures, treating non-stationarity as sequential dynamical regimes in high-dimensional spaces.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Complexity and transition to chaos in coupled Adler-type oscillators.

Physical review. E·2023
Same author

Limiting distributions of continuous-time random walks with superheavy-tailed waiting times.

Physical review. E, Statistical, nonlinear, and soft matter physics·2013
Same author

Asymptotic solutions of decoupled continuous-time random walks with superheavy-tailed waiting time and heavy-tailed jump length distributions.

Physical review. E, Statistical, nonlinear, and soft matter physics·2012
Same author

Continuous-time random walk with a superheavy-tailed distribution of waiting times.

Physical review. E, Statistical, nonlinear, and soft matter physics·2011
Same author

Anomalous biased diffusion in a randomly layered medium.

Physical review. E, Statistical, nonlinear, and soft matter physics·2010
Same author

Directed transport in periodically rocked random sawtooth potentials.

Physical review. E, Statistical, nonlinear, and soft matter physics·2009

Area of Science:

  • Signal Processing
  • Dynamical Systems Theory
  • Acoustics

Background:

  • Deterministic chaotic signals possess unique structures amenable to noise reduction.
  • Human speech exhibits complex, non-stationary characteristics.
  • Existing noise reduction methods may struggle with the inherent complexity of speech.

Purpose of the Study:

  • To adapt a local projective noise reduction scheme for human speech signals.
  • To investigate the applicability of deterministic dynamical system properties to speech.
  • To address challenges posed by speech non-stationarity in noise reduction.

Main Methods:

  • Application of a local projective noise reduction scheme.
  • Analysis of human speech signals within high-dimensional embedding spaces.

Related Experiment Videos

  • Characterization of speech non-stationarity as a sequence of dynamical regimes.

    • Main Results:

    • Successful noise reduction in human speech using the adapted scheme.
    • Demonstration that speech signals exhibit properties similar to deterministic dynamical systems.
    • Resolution of speech non-stationarity into multiple dynamical regimes of moderate complexity.

    Conclusions:

    • The local projective noise reduction scheme is effective for human speech.
    • Speech dynamics can be effectively modeled using principles from deterministic systems.
    • The proposed approach offers a new perspective for speech signal processing and noise reduction.