Artificial Cognition for Early Leaf Disease Detection using Vision Transformers

Abstract

There are many kinds of cassava leaf diseases firmly harm cassava yield, including four main types as followings: Cassava Bacterial Blight (CBB), Cassava Brown Streak Disease (CBSD), Cassava Green Mottle (CGM), and Cassava Mosaic Disease (CMD). In a traditional way, leaf diseases were diagnosed intuitively by farmers. This process is inefficient and unreliable. Several studies have recently relied on deep neural networks for identifying leaf diseases. In this research, we exploit the novel model named Vision Transformer (ViT) in place of a convolution neural network (CNN) for classifying cassava leaf diseases. Experimental results show that this model can obtain competitive accuracy at least 1% higher than popular CNN models (EfficientNet, Resnet50d) on Cassava Leaf Disease Dataset. These results also indicate the potential superiority of the ViT over established methods in analyzing leaf diseases. Next, we quantize the original model and successfully deploy it onto the Edge device named Raspberry Pi 4, which can be attached to a drone that allows farmers to automatically and efficiently detect infected leaves. This result has a significant capability for many future applications in smart agriculture.