A hybrid algorithm for robust acoustic source localization in noisy and reverberant environments

Ramesh Rajagopalan; Timothy Dessonville

doi:10.1117/12.2060758

23 September 2014 A hybrid algorithm for robust acoustic source localization in noisy and reverberant environments

Ramesh Rajagopalan, Timothy Dessonville

Proceedings Volume 9223, Remote Sensing System Engineering V; 92230H (2014) https://doi.org/10.1117/12.2060758
Event: SPIE Optical Engineering + Applications, 2014, San Diego, California, United States

Abstract

Acoustic source localization using microphone arrays is widely used in videoconferencing and surveillance systems. However, it still remains a challenging task to develop efficient algorithms for accurate estimation of source location using distributed data processing. In this work, we propose a new algorithm for efficient localization of a speaker in noisy and reverberant environments such as videoconferencing. We propose a hybrid algorithm that combines generalized cross correlation based phase transform method (GCC-PHAT) and Tabu search to obtain a robust and accurate estimate of the speaker location. The Tabu Search algorithm iteratively improves the time difference of arrival (TDOA) estimate of GCC-PHAT by examining the neighboring solutions until a convergence in the TDOA value is obtained. Experiments were performed based on real world data recorded from a meeting room in the presence of noise such as computer and fans. Our results demonstrate that the proposed hybrid algorithm outperforms GCC-PHAT especially when the noise level is high. This shows the robustness of the proposed algorithm in noisy and realistic videoconferencing systems.

Citation Download Citation

Ramesh Rajagopalan and Timothy Dessonville "A hybrid algorithm for robust acoustic source localization in noisy and reverberant environments", Proc. SPIE 9223, Remote Sensing System Engineering V, 92230H (23 September 2014); https://doi.org/10.1117/12.2060758

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available