ABSTRACT
Even though India is a developing country, many countries worldwide depend on India for food because agriculture is the backbone of India. Nevertheless, farmers cannot receive productivity due to a lack of resources and awareness. So, the proposed paper focuses on smart agriculture to increase the productivity of the tomato crop by identifying the leaf diseases at an early stage by using “Hybrid GAN.” The proposed deep learning classifies healthy and early blight diseases using customized GAN architecture. In general, traditional GANs are used to increase the size of the dataset. However, the proposed model uses the concept of transfer learning and customizes the last layers of neural networks to perform accurate classifications. The model also extracts its features using enhanced autoencoder neural networks. Most of the existing systems apply traditional deep learning techniques with the help of pre-trained models, which are expensive and the farmer cannot afford 262it. The proposed paper simplifies the modification process in the layers and reduces the deployment cost. Usage of deep learning techniques at each stage of design made the model achieve 99.9
