Hyperspectral Sharpening using Scene-adapted Gaussian Mixture Priors

This paper tackles a hyperspectral data fusion problem, using the so-called plug-and-play approach, which combines an ADMM algorithm with a denoiser based on a Gaussian mixture prior. We build upon the concept of scene-adapted prior where, as the name suggests, we learn a model that is targeted to the specific scene being imaged, and show state-of-the-art results on several hyperspectral sharpening experiments. Additionally, we prove that the algorithm is guaranteed to converge.