Computational Mathematics and Scientific Computing Seminar

We develop a new empirical Bayesian inference algorithm for solving a linear inverse problem given multiple measurement vectors (MMV) of under-sampled and noisy observable data.  Specifically, by exploiting the joint sparsity across the multiple measurements in the sparse domain of the underlying signal or image,  we construct a new  support informed sparsity promoting prior. Several applications can be modeled using this framework. Our numerical experiments demonstrate that using this new prior not only improves accuracy of the recovery, but also reduces the uncertainty in the posterior when compared to standard sparsity producing priors.