论文部分内容阅读
Conventional sound localization approaches with small-sized microphone arrays are usually sensitive to noise and reverberation. To deal with the problem, an approach based on expectation maximization algorithm with differential microphone arrays(DMAs) is proposed.Firstly, the parameters of Gaussian mixture model for time-frequency instantaneous direction estimation are estimated through the EM algorithm, and then the direction of each sound source is estimated via time-frequency separation. In order to overcome the weakness of existing time-frequency separation techniques, an improved method, which combines the advantages of both the hard and soft separation methods, is also proposed. The improved time-frequency separation method is shown to be less sensitive to noise and reverberation. Simulation and experimental results demonstrate that the proposed localization approach is superior to its existing counterparts in terms of localization accuracy and robustness.
Conventional sound localization approaches with small-sized microphone arrays are usually sensitive to noise and reverberation. To deal with the problem, an approach based on expectation maximization algorithm with differential microphone arrays (DMAs) is proposed. Firstly, the parameters of Gaussian mixture model for time-frequency instantaneous direction estimation are estimated through the EM algorithm, and then the direction of each sound source is estimated via time-frequency separation. An order to overcome the weakness of existing time-frequency separation techniques, an improved method, which combines the The advantages of both the hard and soft separation methods, is also proposed. The improved time-frequency separation method is shown to be less sensitive to noise and reverberation. Simulation and experimental results demonstrate that the proposed localization approach is superior to its existing counterparts in terms of localization accuracy and robustness.