Support data for paper "Revisiting remote sensing cross-sensor Single Image Super-Resolution: the overlooked impact of geometric and radiometric distortion"

Contents

This dataset contains support data for the following paper:

Julien Michel, Ekaterina Kalinicheva, Jordi Inglada. Revisiting remote sensing cross-sensor Single Image Super-Resolution: the overlooked impact of geometric and radiometric distortion. 2024. ⟨hal-04723225⟩

It contains:

dataset.tar.gz : Datasets wsx4 and wsx2 generated from the Worldstrat dataset,

pretrained_registration_models.tar.gz : All pre-trained registration models required to replay the experiments,

trained_models.tar.gz : all trained models from the experiments,

onnx_exported_models.tar.gz : all trained models from the experiments, exported to ONNX and ready to be used with the sentinel2_superresolution tool for testing on any Sentinel-2 image.

Licences

File dataset.target.gz and all files contained in the archive are covered by the CC-BY-NC 4.0 licence (pertaining the original licence from Worldstrat)

All other files are covered by the CC-BY-4.0 licence.