Multi-heat keypoint incorporation in deep learning model to tropical cyclone centering and intensity classifying from geostationary satellite images

Abstract Hydrometeorological forecasting and early warning involve many hazardous elements, with the estimation of intensity and center location of tropical cyclones (TCs) being key. This paper proposes a new multitask deep learning model with attention gate mechanisms to work with satellite images...

Full description

Saved in:
Bibliographic Details
Main Authors: Thanh-Ha Do, Son-The Phan, Duc-Tien Du, Dinh-Quan Dang, Khanh-Hung Mai, Lars R. Hole
Format: Article
Language:English
Published: Nature Portfolio 2025-07-01
Series:Scientific Reports
Online Access:https://doi.org/10.1038/s41598-025-12733-w
Tags: Add Tag
No Tags, Be the first to tag this record!