Defects Detection Dataset

v3

2024-02-08 10:01pm

Generated on Feb 8, 2024

Dataset Split

Train Set 93%
63Images
Valid Set 3%
2Images
Test Set 4%
3Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 800x600

Augmentations

Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise
Crop: 0% Minimum Zoom, 60% Maximum Zoom
Grayscale: Apply to 25% of images
Noise: Up to 1.96% of pixels
Cutout: 15 boxes with 10% size each