GlobeNet: Convolutional Neural Networks for Typhoon Eye Tracking from Remote Sensing Imagery

TL;DR

This paper introduces GlobeNet, a convolutional neural network approach utilizing high-resolution satellite imagery to improve typhoon eye tracking, demonstrating promising results in predicting typhoon coordinates.

Contribution

The paper presents a novel CNN-based method for typhoon tracking using multichannel satellite images, highlighting specific model combinations and activation policies for better predictions.

Findings

Effective prediction of typhoon coordinates in the northeastern hemisphere

Use of multichannel satellite imagery enhances atmospheric understanding

Model configurations significantly impact prediction accuracy

Abstract

Advances in remote sensing technologies have made it possible to use high-resolution visual data for weather observation and forecasting tasks. We propose the use of multi-layer neural networks for understanding complex atmospheric dynamics based on multichannel satellite images. The capability of our model was evaluated by using a linear regression task for single typhoon coordinates prediction. A specific combination of models and different activation policies enabled us to obtain an interesting prediction result in the northeastern hemisphere (ENH).