This talk has been automatically
captioned. You can remove these by pressing CC on the video toolbar.
Name: John Nolan
Talk Title: Robust Sparse Reconstruction
Talk Abstract: Traditional compressive sensing (CS) assumes light-tailed models for the underlying signal and/or noise. This assumption is not met in the case of highly impulsive environments, where non-Gaussian processes arise. In this case, traditional sparse reconstruction methods perform poorly, since they are incapable of suppressing the effects of heavy-tailed sampling noise. This talk will describe the use of heavy-tailed stable distributions to design a robust algorithm for sparse signal reconstruction from linear random measurements corrupted by infinite-variance additive noise. We demonstrate the improved reconstruction performance of the proposed algorithm when compared against standard CS techniques for a broad range of impulsive environments.
This
talk is an invited talk at EVA 2021.