YoloNet: Hybrid YOLO11 and EfficientNet for Lumpy Skin Disease Detection

The Lumpy Skin Disease (LSD) is a very infectious and contagious disease which affects the lives of countless cattle each year. Machine/Deep Learning models like CNNs effectively detect this disease through images and track the progression of this disease in infected cattle. Various works have been done on this problem, using models such as MobileNetV2, ResNet50, Gradient Boost, Xception, Inception V3, and DenseNet121. This paper proposes a hybrid model, YoloNet, to detect the LSD virus in Cattle, using YOLO11 and EfficientNet-B0. The YoloNet model takes a pre-processed image and performs feature extraction using YOLO11 to generate a tensor. The generated tensor is given as input to EfficientNet-B0, which converts it to a feature map and classifies the image. The analysis of the YoloNet has been done on a dataset with 5 different classes of cattle, with 8014 total images. The accuracy of the YoloNet model, being 97.97<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup>, is significantly better than other pre-existing state-of-the-art models.