( 03 ) Deep Learning
Flood Prediction
Machine-learning flood prediction models enable large-scale, high-resolution inundation forecasting that conventional numerical simulation could not achieve.

Real-time Flood Prediction with Deep Learning
深層学習と次元圧縮技術の融合による、リアルタイム浸水域予測手法の構築
Real-time Flood Prediction with Deep Learning
Developing a surrogate model that combines dimensionality reduction and deep learning to predict flood inundation areas in seconds while maintaining accuracy comparable to physics-based simulation.
#Deep Learning#River Flooding#Real-time Prediction

Underwater Object Classification with Deep Learning
深層学習に基づいた水中物体分類手法の構築
Underwater Object Classification with Deep Learning
Using deep learning to classify and colorize point clouds by structure, making underwater survey data - which lacks color information - easier to interpret.
#Deep Learning#Point Cloud#Underwater Survey

Roughness Estimation from Point Clouds
点群を用いた粗度算出手法の構築
Roughness Estimation from Point Clouds
Estimating roughness coefficients from terrain features captured in LiDAR point clouds to improve the accuracy of flood inundation simulation.
#Point Cloud#Roughness Coefficient#Flood Simulation