/RDD-YOLOv5

Primary LanguagePythonGNU General Public License v3.0GPL-3.0

RDD-YOLOv5

This program is an improved-YOLOv5 version, used for road defect detection. The original version is https://github.com/ultralytics..

The paper is RDD-YOLOv5: Road Defect Detection Algorithm with Self-Attention Based on Unmanned Aerial Vehicle Inspection. Discussions and suggestions are welcome.

The architecture can seen in the figure: image

Relative network configuiration is in the https://github.com/arandinglv/RDD-YOLOv5/blob/main/models/yolov5m_2evc_swin_gelu.yaml

The Road Defect Dataset is uploaded on BaiduNetDisk:

URL: https://pan.baidu.com/s/1naRqVZ7qoZMLrqXENdKBKg?pwd=5tjh

CODE: 5tjh

More information about dataset is here: https://github.com/arandinglv/RoadDefectDataset

Noted that files about visualization of P-R-PR are packed in the https://github.com/arandinglv/RDD-YOLOv5/tree/main/p_r_pr_f1_painter, more information can be referred the PDF file.