August 2, 2025

EADI-YOLO: a lightweight and efficient model for rail surface defect detection

View Full Paper

Ask AI

Bookmark

Authors

Discussion

EADI-YOLO demonstrates improved detection accuracy and speed in rail surface defect detection, highlighting its efficiency.

EADI-YOLO achieves a mAP@0.5 of 90.3% for rail surface defect detection, indicating high accuracy.
Experimental results show a 2.8% and 6.7% improvement in detection accuracy compared to YOLOv10n.
The model replaces standard convolutions with the AKConv module to enhance multi-scale defect perception.
Generalization experiments confirm EADI-YOLO's balance of accuracy, speed, and model complexity across various datasets.

Ma et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark