AI Insight
This study introduces KT-YOLO, a modified object detection model based on YOLOv8n, designed to identify four behavioral categories (standing, lying, eating, drinking) in densely packed Hu sheep under intensive farming conditions. The model incorporates a Kernel-Team Fusion (KTF) method using four different convolution kernel sizes to better capture multi-scale features and handle occlusion, alongside a SlideLoss function to compensate for class imbalance in the training dataset. Tested on the newly created Hu Sheep Behavior Dataset (HSBD, 280 images, 6,766 sheep instances), KT-YOLO achieved a mAP50 of 86.4%, outperforming the baseline YOLOv8n by 6.3 percentage points.
Why it matters
Automated behavioral monitoring in livestock farming can help farmers detect early signs of illness, stress, or abnormal activity without continuous human observation, potentially improving animal welfare and operational efficiency in intensive sheep farming.
Understand the Science
by Suoxiang Zhang, Hongrui Chang, Zhonghong Wu, Guoxin Wu, Ronghua Ji
Computer vision has been extensively applied to sheep behavior detection in recent years. However, the dense distribution of Hu sheep poses detection challenges, while imbalanced behavioral categories in datasets affect classification accuracy for detection tasks in intensive farming scenarios, resulting in high misclassification rates. Current models often rely on over-parameterization to achieve satisfactory detection performance, which increases computational burden and limits practical deployment. To address these challenges, this study introduces the Hu Sheep Behavior Dataset (HSBD), specifically designed for intensive farming environments. The dataset comprises 280 images capturing four behaviors across 6,766 Hu sheep: standing, lying, eating, and drinking. Building upon this foundation, we developed the KT-YOLO model, which utilizes a novel Kernel-Team Fusion (KTF) method to enhance the YOLOv8n detection framework. By employing four different convolution kernel sizes, this method effectively captures multi-scale features and addresses Hu sheep occlusion challenges. To mitigate accuracy degradation caused by dataset imbalance, KT-YOLO incorporates a SlideLoss function during classification, effectively addressing this challenge. Comparative experiments demonstrate that KT-YOLO achieved a mean Average Precision (mAP50) of 86.4%, representing a 6.3 percentage point improvement over YOLOv8n, with SlideLoss contributing an additional 1 percentage point improvement. Further comparison with YOLOv13n demonstrates KT-YOLO’s superior performance in dense Hu sheep behavior detection. By introducing HSBD and developing the innovative KT-YOLO, this study significantly enhances both accuracy and efficiency of dense Hu sheep behavior detection, demonstrating the potential and practical value of deep learning technologies in intensive farming environments.
Source: KT-YOLO: A multi-convolution kernel collaboration model for dense Hu sheep behavior detection