Classification of Cassava Leaf Diseases Using a Pre-trained EfficientNet Model

Cassava is one of the staple crops in most tropical and subtropical countries and said to supply the vital food for millions of people. However, the production of this crop is substantially challenged by many leaf diseases, such as the Cassava Mosaic Disease (CMD), the Cassava Brown Streak Disease (CBSD), and the Cassava Bacterial Blight (CBB). In contrast, the traditional methods of disease detection are laborious, time-consuming, and prone to errors due to human intervention in their very nature. This paper works on a pre-trained EfficientNet model for automated cassava leaf disease classification. The work takes advantage of the latest deep-learning techniques to boost the efficiency and accuracy of the process of classifying cassava leaves. As we can see the model gave excellent performance by achieving maximum accuracy at 92.83% with a minimum loss of 0.2019 measured at epoch 16.