Deep blind super-resolution for hyperspectral images

Published: 01 Jan 2025, Last Modified: 29 Sept 2024Pattern Recognit. 2025EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•A novel blind super-resolution algorithm for the single hyperspectral image is proposed for the first time, effectively addressing the issue of unknown degradation in hyperspectral images.•We propose a blur kernel estimation module and a deblurring module to obtain the clean hyperspectral image without blur.•We design the branch and fusion module to exploit the spatial-spectral information from both the local and global perspectives to further remove degradation.•Experimental results validate that the proposed method has superior performance and better generalization capability.
Loading

OpenReview is a long-term project to advance science through improved peer review with legal nonprofit status. We gratefully acknowledge the support of the OpenReview Sponsors. © 2025 OpenReview