A Large Scale Open-Source Image and Video Dataset for Robust Wildfire Detection and Classification
Researchers have introduced the Global Wildfire Prevention Dataset (GWFP), a new open-source collection of wildfire images and videos aimed at advancing early detection and classification research. The dataset features diverse wildfire scenes, including flames, smoke, and challenging environmental conditions, collected globally. It is benchmarked with various convolutional and transformer architectures to assess robustness and generalization capabilities, with a novel HTE-ResNet model showing promising results. AI
IMPACT This dataset could significantly improve the accuracy and speed of AI-powered wildfire detection systems, leading to better early response and mitigation efforts.