Reimagining Disease Detection in Cassava Leaves with Advanced AI Techniques

In the Philippines, there are many diseases that seriously harm the growth and production of cassava, which is the most important crop. Since people rely on cassava for food, it’s very important to find and deal with diseases early before they get worse. The methods used for finding diseases rely on the expert knowledge of related products and the experience of the people involved. This work presents an automatic system that uses existing models of neural networks and deep learning, based on methods from convolutional neural networks, to manage unclear data for recognizing diseases. In this study, we used a model called EfficientnetB3 to identify diseases in cassava plants with unclear data. The model had 10,791,220 parameters that could be adjusted during training, and we used a method called transfer learning to help with this process. Our method for finding crop diseases can help automatically collect information to identify diseases and stop crop loss.