A track before detect approach for sequential Bayesian tracking of multiple speech sources

This paper describes a novel multiple acoustic source tracking method based on track before detect paradigm. Multiple particle filters are used to represent the state of all sources. Sources are detected and removed using a likelihood ratio obtained from particle weights. The weights are obtained by evaluating the likelihood of microphone pair phase difference. Tracking performance from recorded data with rich sequences of speech is presented using multiple object tracking metrics. Results show that the proposed method can detect and track multiple temporally overlapping speech sources as well as switching talkers even in weak signal-to-noise ratios.

[1]  Alessio Brutti,et al.  A sequential Monte Carlo approach for tracking of overlapping acoustic sources , 2009, 2009 17th European Signal Processing Conference.

[2]  Jean Rouat,et al.  Robust localization and tracking of simultaneous moving sound sources using beamforming and particle filtering , 2007, Robotics Auton. Syst..

[3]  Rainer Stiefelhagen,et al.  Evaluating Multiple Object Tracking Performance: The CLEAR MOT Metrics , 2008, EURASIP J. Image Video Process..

[4]  Eric A. Lehmann,et al.  Particle filtering methods for acoustic source localisation and tracking , 2004 .

[5]  M. Fallon,et al.  Multi Target Acoustic Source Tracking with an Unknown and Time Varying Number of Targets , 2008, 2008 Hands-Free Speech Communication and Microphone Arrays.

[6]  Benoît Champagne,et al.  Performance of time-delay estimation in the presence of room reverberation , 1996, IEEE Trans. Speech Audio Process..

[7]  Neil J. Gordon,et al.  A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking , 2002, IEEE Trans. Signal Process..

[8]  Guillaume Lathoud,et al.  Spatio-Temporal Analysis of Spontaneous Speech with Microphone Arrays , 2006 .

[9]  W. Ritter,et al.  Detection and Tracking of Multiple Pedestrians in Automotive Applications , 2007, 2007 IEEE Intelligent Vehicles Symposium.

[10]  J. N. Driessen,et al.  Particle filter based detection for tracking , 2001, Proceedings of the 2001 American Control Conference. (Cat. No.01CH37148).

[11]  Carlos Busso,et al.  Multimodal Meeting Monitoring: Improvements on Speaker Tracking and Segmentation through a Modified Mixture Particle Filter , 2007, 2007 IEEE 9th Workshop on Multimedia Signal Processing.

[12]  Joseph H. DiBiase.  A High-Accuracy, Low-Latency Technique for Talker Localization in Reverberant Environments Using Microphone Arrays , 2000 .