A latent rectified-flow model jointly forecasts nine hours of tropical cyclone satellite imagery and ERA5 winds, temperature, and pressure in one pass, beating a reproduced diffusion baseline on image quality while using far fewer sampling steps.
Improving tropical cyclone forecasting with video diffusion models
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Tropical Cyclone Forecasting via Latent Rectified Flow using Satellite Imagery and Atmospheric Fields
A latent rectified-flow model jointly forecasts nine hours of tropical cyclone satellite imagery and ERA5 winds, temperature, and pressure in one pass, beating a reproduced diffusion baseline on image quality while using far fewer sampling steps.