Chicken Disease Diagnosis Model Using YOLOv8 Object Detection Algorithm with SE-Attention Mechanism

Traditional chicken disease diagnosis methods rely on manual observation and experiential judgment, which are inefficient and susceptible to subjective factors. This paper aims to construct a chicken disease diagnostic model with deep learning algorithm for enhancing the accuracy and efficiency. Firstly, this paper proposes a chicken disease diagnostic model based on YOLOv8 algorithm. An SE-attention mechanism is designed for YOLOv8 structure to improve the detection accuracy. The SE-Attention based YOLOv8 detection model can identify and classify diseases by analyzing the feces images of chickens. Experiments on chicken disease diagnosis are performed to validate the proposed model's feasibility and effectiveness. Ablation study is constructed to validate the advantages of the SE-attention mechanism.