Adaptive Hybrid Optimizer based Framework for Lumpy Skin Disease Identification

Lumpy Skin Disease (LSD) is a contagious viral infection that significantly deteriorates livestock health. Early and precise identification is crucial. This paper proposes a hybrid deep learning-based approach called LUMPNet for the early detection of LSD, utilizing image data to detect and classify skin nodules. LUMPNet uses YOLOv11, EfficientNet-based CNN classifier, and a novel adaptive hybrid optimizer. Results indicate 99% LSD detection training accuracy and 98% validation accuracy, outperforming existing schemes.