Related Experiment Video
Updated: Aug 14, 2026
03:31
End-To-End Deep Neural Network for Salient Object Detection in Complex Environments
Published on: December 15, 2023
GMS-YOLO11n: A Sheep Detection Model for Challenging Fixed-View Farm Conditions Integrating Spatially Gated
Wenbo Yu1,2, Ruoya Xie1,2, Yongqi Liu1,2
1College of Mechanical and Electrical Engineering, Inner Mongolia Agricultural University, Hohhot 010018, China.
Animals : an Open Access Journal From MDPI
|August 13, 2026
Summary
This study introduces GMS-YOLO11n, an enhanced object detection model for sheep monitoring in intelligent farming. It significantly improves sheep detection accuracy, especially under challenging conditions like low light and occlusion.
Area of Science:
- Computer Vision
- Artificial Intelligence
- Livestock Management
Background:
- Accurate sheep detection is crucial for intelligent livestock farming operations.
- Challenges include occlusion, scale variation, low light, and background interference, leading to detection errors.
- Existing methods struggle with these complex environmental factors.
Purpose of the Study:
- To develop an improved object detection model, GMS-YOLO11n, for fixed-view sheep monitoring.
- To enhance detection accuracy and localization under various challenging conditions.
- To provide a robust solution for intelligent sheep farming.
Main Methods:
- Proposed GMS-YOLO11n, an improved YOLO11n-based detector.
- Introduced a spatially gated bottleneck convolution module to enhance local structural cues.
- Implemented a multi-scale attention fusion module for integrating semantic and detailed features.
- Evaluated on a custom dataset of 3531 Small-tailed Han sheep images (8167 instances).
Main Results:
- GMS-YOLO11n achieved high performance: 93.62% precision, 91.03% recall, 92.30% F1-score, 96.53% mAP@0.5, and 76.32% mAP@0.5:0.95.
- Demonstrated significant improvements over the YOLO11n baseline, particularly in mAP@0.5:0.95 (+6.92%).
- Showed substantial gains under low-light (+9.83%) and occlusion (+7.62%) conditions.
Conclusions:
- GMS-YOLO11n effectively improves sheep detection performance in challenging visual conditions.
- The model shows promise for real-world applications in intelligent sheep farming.
- Further external validation is recommended for broader generalization.