A Disaster Response Model for Tropical Storm Damage Detection in Satellite Imagery

Storms cause massive destruction around the world and are increasing in frequency. The outcome of this project saves lives and resources by streamlining disaster response and helping governments and communities plan effectively. Our model is trained using detailed cyclone Hurricane Maria imagery, with a Mean Average Precision (mAP) score of 0.46 and distinguishes between damaged and undamaged buildings. Our tool can revolutionize disaster response, saving lives and reducing costs.