Abstract: This work explores capabilities of the pre-trained CLIP vision-language model to identify satellite images affected by clouds. Several approaches to using the model to perform cloud presence ...
Abstract: Spectral pixels are often a mixture of the pure spectra of the materials, called endmembers, due to the low spatial resolution of hyperspectral sensors, double scattering, and intimate ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback