Technical Introduction

HARMONIC-TEMPORAL CLUSTERING OF SPEECH FOR SINGLE AND

MULTIPLE F0 CONTOUR ESTIMATION IN NOISY ENVIRONMENTS

Jonathan Le Roux, Kameoka Hirokazu, Nobutaka Ono, Alain de Cheveigné, Shigeki Sagayama

Sagayama/Ono Lab., The University of Tokyo.

1. Abstract

2. Motivation and Approach

3. Derivation of the Model

4. Optimization of the Model Parameters

5. Experimental Evaluation (1): Single-Speaker F0 Estimation in Clean Environment

6. Experimental Evaluation (2): Single-Speaker F0 estimation in Noisy Environments

7. Experimental Evaluation (3): Multiple F0 Estimation of Co-Channel Concurrent Speech

8. Conclusion and Bibliography

This work was presented at the 32nd IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), which was held at the Hawai‘i Convention Center in Honolulu, April 15 - 20, 2007.

Acknowledgment: Part of the expenses for the presentation of this work at ICASSP 2007 was supported by a grant from The Telecommunications Advancement Foundation (TAF, Japan).

Back to the overview page (English)

Back to the overview page (Japanese)

Sagayama/Ono Lab. HP (English)

Sagayama/Ono Lab. HP (Japanese)