Skip to content

Latest commit

 

History

History
32 lines (25 loc) · 2.22 KB

README.md

File metadata and controls

32 lines (25 loc) · 2.22 KB

Monitoring war damage using satellite data and artificial intelligence

Description

War-related urban destruction constitutes a significant global concern, affecting the environment, social fabric, and urban planning. The impacts on complex urban morphologies during conflicts, characterized by different levels of destruction, remain inadequately understood. Here we present how deep learning, integrated with very high-resolution satellite imagery, enables near real-time detection of explosions and evaluation of different building damage levels in an active conflict zones.

Requirements

Python 3.7+ Pytorch 1.7.1 torchvision 0.8.2 Opencv 4.5.5 CUDA Toolkit 10.1 Python-SNAPPY 8.0

🔭 Baselines

  • 📖 📖 📖 Baseline [here]
  • 📖 📖 📖 AFDE-Net [here]
  • 📖 📖 📖 U-Net [here]
  • 📖 📖 📖 MS4D-Net [here]

References