Haiyan Song
Papers
3
Total Citations
43
H-Index
3
About
Dr. Haiyan Song is a leading researcher in agricultural artificial intelligence and precision robotics, with a focus on deep learning for crop detection and segmentation in complex natural environments. Her work centers on developing lightweight, high-accuracy computer vision models—such as YOLOv8-ECFS and AHG-YOLO—to enable real-time, multi-category detection of weeds in soybean fields and occluded pear fruits in orchards, directly supporting autonomous harvesting and weeding robots. Dr. Song has also advanced instance segmentation techniques, proposing an improved Mask R-CNN model incorporating Swin-Transformer for precise segmentation of Zanthoxylum bungeanum clusters, a critical step for robotic picking in unstructured settings. Her most cited paper, “YOLOv8-ECFS: A lightweight model for weed species detection in soybean fields” (2024), has already garnered 29 citations, reflecting its immediate impact on efficient, field-deployable AI. With additional influential works on occluded fruit detection and crop cluster segmentation, Dr. Song’s contributions are shaping the next generation of intelligent agricultural machinery, bridging the gap between cutting-edge deep learning and practical, real-world farming challenges.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3