Related Experiment Video
Updated: Mar 6, 2026
03:31
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
Published on: December 15, 2023
1.2K
Rice pest detection via multi-scale edge network and wavelet attention enhancement
1School of Software Engineering, Jiangxi University of Science and Technology, Nanchang, China.
Frontiers in Plant Science
|March 5, 2026
Summary
This study introduces BEAM-YOLO, an advanced AI model for detecting rice pests. It significantly improves accuracy in identifying small and similar-looking pests in complex agricultural fields.
Area of Science:
- Agricultural Science
- Computer Vision
- Artificial Intelligence
Background:
- Rice pest detection is challenging due to small targets, similar appearances, and complex backgrounds.
- Existing methods struggle with accurate identification in real-world field conditions.
Purpose of the Study:
- To develop an efficient and accurate deep learning model for rice pest detection.
- To address limitations in recognizing small, morphologically similar pests against cluttered backgrounds.
Main Methods:
- Proposed BEAM-YOLO (Bi-branch Edge Attention Multi-scale YOLO) model.
- Developed the JRICE-PD dataset with 11 rice pests (4,565 images).
- Introduced four novel modules: MEN, BAFE, EM-BFPN, and SCAU for enhanced feature extraction and small target detection.
Main Results:
- BEAM-YOLO achieved 86.6% mAP@50 and 72.7% mAP@50-95.
- Outperformed YOLOv11 by 3.3% (mAP@50) and 3.0% (mAP@50-95).
- Maintained efficient computational overhead and parameter count.
Conclusions:
- BEAM-YOLO offers a robust solution for intelligent agricultural pest monitoring.
- The model provides reliable algorithmic support for advancing precision agriculture.
- This research enhances automated systems for crop protection and yield optimization.