Dataset Versions
v1
2023-06-22 6:24pm
Generated on Jun 22, 2023
2612 Total Images
View All Images
Dataset Split
Train Set 92%
2391Images
Valid Set 6%
151Images
Test Set 3%
70Images
Preprocessing
Auto-Orient: Applied
Resize: Stretch to 640x640
Augmentations
Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Rotation: Between -15° and +15°
Shear: ±13° Horizontal, ±23° Vertical
Grayscale: Apply to 25% of images
Blur: Up to 4px
Noise: Up to 20% of pixels
Cutout: 5 boxes with 18% size each