AbstractHail, a highly destructive weather phenomenon, necessitates critical identification and forecasting for the protection of human lives and properties. The identification and forecasting of hail are vital for ensuring human safety and safeguarding assets. This research proposes a deep learning algorithm named Dual Attention Module EfficientNet (DAM-EfficientNet), based on EfficientNet, for detecting hail weather conditions. DAM-EfficientNet was evaluated using FY-4A satellite imagery and real hail fall records, achieving an accuracy of 98...
