Disease Detection of Solanum Lycopersicum using SqueezeNet Deep Learning Architecture

The productivity of a nation's food supply hinges on the health of its crops, and plant diseases can have a severe impact on food production. Early detection of these diseases is vital, and this can be achieved through analyzing images of plant leaves. Advances in deep learning have notably enhanced the detection of plant diseases. This study introduces SqueezeNet for classifying diseases in Solanum Lycopersicum (tomato plants) and compares its performance to other advanced deep learning models. The research involved classifying leaves into ten disease categories, using the augmented plant Village dataset for training, validation, and testing via transfer learning. The results showed that SqueezeNet1_0 and SqueezeNet1_1 achieved the highest test accuracies of 99% and 99.8%, respectively, outperforming other models like VGG16, GoogLeNet, EfficientNet B0, and ResNet50.