An Enhanced YOLOv4 Model With Self-Dependent Attentive Fusion and Component Randomized Mosaic Augmentation for Metal Surface Defect Detection.
Chenglong WangZiran ZhouZhiming ChenPublished in: IEEE Access (2022)
Keyphrases
- experimental data
- mathematical model
- high level
- artificial neural networks
- computational model
- defect detection
- object model
- information fusion
- management system
- image registration
- fusion model
- image fusion
- computer vision
- probability distribution
- probabilistic model
- objective function
- image segmentation
- feature selection