Hyperspectral super-resolution via coupled tensor ring factorization

Hyperspectral super-resolution (HSR) fuses a low-resolution hyperspectral image (HSI) and high-resolution multispectral (MSI) to obtain HSI (HR-HSI). In this paper, we propose new model called coupled tensor ring factorization (CTRF) for HSR. The proposed CTRF approach simultaneously learns the core tensors of HR-HSI from pair MSI. can separately exploit low-rank property each class (Section 3.3), which has not been explored in previous models. Meanwhile, inherits simple representation matrix/canonical polyadic flexible exploration Tucker factorization. We further introduce spectral nuclear norm regularization explore global property. experiments demonstrated advantage regularized compared matrix/tensor deep learning methods.