Real-Time Wildfire Detection Using ResNet50V2

This research provides a convolutional neural network (CNN) model for feature recognition between real and bogus ones. The model was trained across ten epochs with a batch size of 64 and a learning rate of 0.001. The design includes many convolutional layers after pooling layers and activation functions (ReLU) to aid in feature extraction. Flexibility of the model was improved by use of data augmentation methodologies. The training approach includes 2,278 photographs for validation and a set of 6,799 images for training. On the validation set, the model demonstrated to be 98.5% accurate, thereby proving its capacity to distinguish between true and spurious attributes. The end loss evaluated at 0.05 reflects a well-fitted model since the loss throughout training fell considerably. These results illustrate how successfully CNNs might manage face classification.