New techniques for solving inverse problems using regularization with sparsity promoting metrics promise improved results for some specific situations. However, these techniques have very high computational costs. Recently, the Kronecker Array Transform (KAT) was developed to provide significant reduction of the computational cost for array processing, with the single restriction that the microphones should be distributed in a separable geometry. This postdoctoral project aims to design and build a microphones array with separable geometry. After verifying the calibration of the arrangement, measurements shall be performed to allow a comparison between these experimental results and those obtained via simulation.
News published in Agência FAPESP Newsletter about the scholarship: