This paper introduces an optimization problem to estimate the covariance matrix from multiple subsets of compressive measurements using random projection matrices. The proposed optimization is tested with computational simulations for the DD-CASSI and SSCSI optical architectures.
