Compact Multi-level-prior Tensor Representation for Hyperspectral Image Super-resolution
The model introduces a novel hyperspectral super-resolution method that compactly represents complex, multi-level priors of images using a tensor framework. It achieves this by decoupling spectral and spatial information through block term decomposition and employing a specialized non-convex total variation constraint. This allows for an efficient iterative optimization process with guaranteed convergence.