Quasi-Monte Carlo Filtering for Speaker Tracking
基于拟蒙特卡洛滤波的说话人跟踪方法

A mean shift quasi-Monte Carlo (MS-QMC) method is proposed for speaker tracking. To explore the state space more efficiently, deterministic samplers are used instead of random draws according to a quasi-Monte Carlo integration rule in the new method. Furthermore, a mean shift procedure is applied to move particles toward the modes of the posterior, leading to a more effective allocation of particles thereupon fewer particles are needed. Simulation results show that compared with the traditional particle filter, both speaker tracking accuracy and convergent rate of the proposed method are improved.