Spatio-Temporal Diffusion Model for Satellite Imagery

Spatio-Temporal Diffusion Model for Satellite Imagery
Prathap Kashyap, Alireza Javanmardi, Pragati Jaiswal, Alain Pagani, Gerd Reis
In: Proceedings of. International Conference on Remote Sensing and Geoinformation of Environment (RSCy-2025), SPIE, 2025.

Abstract:

Generative AI models have proven exceptionally useful for learning representations of the real world and creating novel data. The aim of this work is therefore the use of open-source data from satellites such as the Landsat and Sentinel series to learn the progression of landscapes over time under various influences and predict their potential future appearances using a diffusion model.