Heterogeneous Bayesian compressive sensing for sparse signal recovery

This study focuses on the issue of sparse signal recovery with sparse Bayesian learning in the context of a heterogeneous noise model, called by the heterogeneous Bayesian compressive sensing. The main contribution is to exploit the capability of noise variance learning in performance improvement and applicability enhancement. Experimental results on synthetic and real-world data demonstrate that heterogeneous Bayesian compressive sensing has superior performance in terms of accuracy and sparsity for both homogeneous and heterogeneous noise scenarios.