Dataset Versions

v4

2024-04-01 9:48am

Generated on Apr 1, 2024

Dataset Split

Train Set 88%
147Images
Valid Set 8%
14Images
Test Set 4%
7Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 700x700

Augmentations

Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Rotation: Between -13° and +13°
Shear: ±10° Horizontal, ±10° Vertical
Noise: Up to 0.18% of pixels