About this document
DiffusionSat-A Generative Foundation Model For Satellite Imagery by Purbandini is a document available to read on EtoBox.
The paper introduces DiffusionSat, a generative foundation model specifically designed for satellite imagery, addressing the limitations of existing diffusion models that are not tailored for remote sensing data. By incorporating metadata such as geolocation and timestamps, DiffusionSat achieves state-of-the-art performance in tasks like super-resolution, temporal generation, and in-painting. This model represents the first large-scale generative foundation model for satellite imagery, trained on high-resol
- Author
- Purbandini
- Language
- EN