Precise Pitch Detection in Noise Environments

DOI

10.2991/icca-16.2016.73How to use a DOI?

Keywords

Pitch detection, Spectral tracking, Pectinate filter, Ponicare pitch map

Abstract

It is still challenging to detect speech pitch precisely in various noise environments at low SNR (signal to noise ratio). Voiced speech harmonics can be clearly found in noisy speech. Some of the harmonic frequencies with higher energy were estimated by spectral tracking method, a pectinate filter was designed according to those harmonic frequencies to filter the noisy speech, and the ponicaré pitch detection method was used to detect pitch from the filtered speech. The error is less than 2Hz and accuracy ratio is more than 90%, which exceeds parallel methods’ by 20%, under the experimental condition of 5 noisy speech and -10dB SNR. Experiments demonstrate the effectiveness of the pitch detection method in various noise environments at low-SNR.

Copyright

© 2016, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Tianwei Li
AU  - Zhengyou Li
AU  - Qian Huang
AU  - Jiao Guo
PY  - 2016/01
DA  - 2016/01
TI  - Precise Pitch Detection in Noise Environments
BT  - Proceedings of the 2016 International Conference on Intelligent Control and Computer Application
PB  - Atlantis Press
SP  - 308
EP  - 311
SN  - 2352-538X
UR  - https://doi.org/10.2991/icca-16.2016.73
DO  - 10.2991/icca-16.2016.73
ID  - Li2016/01
ER  -