Abstract
In this paper, a study addressing the task of tracking multiple concurrent speakers in reverberant conditions is presented. Since both past and future observations can contribute to the current location estimate, we propose a forward-backward approach, which improves tracking accuracy by introducing near-future data to the estimator, in the cost of an additional short latency. Unlike classical target tracking, we apply a non-Bayesian approach, which does not make assumptions with respect to the target trajectories, except for assuming a realistic change in the parameters due to natural behaviour. The proposed method is based on the recursive expectation-maximization (REM) approach. The new method is dubbed forward-backward recursive expectation-maximization (FB-REM). The performance is demonstrated using an experimental study, where the tested scenarios involve both simulated and recorded signals, with typical reverberation levels and multiple moving sources. It is shown that the proposed algorithm outperforms the regular common causal (REM).
Highlights
The task of multiple target tracking has significant importance in civil, military and surveillance applications such as improving beamforming accuracy in speech enhancement applications, e.g. speech separation, indoor robotic assistance, and automatic steering of cameras [1,2,3,4]
We propose a new tracking mechanism and use it to modify the recursive distributed expectationmaximization (RDEM) [51], resulting in the tracking forward-backward recursive expectation-maximization (TFB-REM), which is a non-Bayesian algorithm
4.2 recursive distributed expectation-maximization (RDEM) applied in the forward direction In [44] and [51], the tracking forward-recursive expectation-maximization (TF-REM) was derived for the general algorithm in (11) in detail, and only the resulting formulae are given
Summary
The task of multiple target tracking (or dynamic localization) has significant importance in civil, military and surveillance applications such as improving beamforming accuracy in speech enhancement applications, e.g. speech separation, indoor robotic assistance, and automatic steering of cameras [1,2,3,4]. A further study of the recursive expectation-maximization (REM) approach appeared in [42] for the problem of DOA estimation, using TREM and another recursive algorithm suggested by the authors. We propose a new tracking mechanism and use it to modify the recursive distributed expectationmaximization (RDEM) [51], resulting in the tracking forward-backward recursive expectation-maximization (TFB-REM), which is a non-Bayesian algorithm.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: EURASIP Journal on Audio, Speech, and Music Processing
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
