Multipitch Tracking for Noisy and Reverberant Speech

Abstract

Multipitch tracking in real environments is critical for speech signal processing. Determining pitch in reverberant and noisy speech is a particularly challenging task. In this paper, we propose a robust algorithm for multipitch tracking in the presence of both background noise and room reverberation. An auditory front-end and a new channel selection method are utilized to extract periodicity features. We derive the conditional probability given each pitch state, which estimates the likelihood of the observed periodicity features given pitch candidates. A hidden Markov model integrates these probabilities and searches for the best pitch state sequence. Our algorithm can reliably detect single and double pitch contours in noisy and reverberant conditions. Quantitative evaluations show that our approach significantly outperforms existing ones, particularly in reverberant conditions.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 01, 2009
Accession Number
AD1001191

Entities

People

  • DeLiang Wang
  • Zhaozhang Jin

Organizations

  • Ohio State University

Tags

Communities of Interest

  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Acoustics
  • Algorithms
  • Broadband
  • Cognitive Science
  • Computational Science
  • Computer Science
  • Computers
  • Detection
  • Engineering
  • Frequency
  • Hidden Markov Models
  • Language
  • Markov Models
  • Probability
  • Probability Distributions
  • Signal Processing
  • White Noise

Fields of Study

  • Computer science
  • Engineering

Readers

  • Auditory Neuroscience/Auditory Physiology.
  • Distributed Systems and Data Platform Development
  • Robotics and Automation.