v1

2024-07-12 4:49am

Generated on Jul 12, 2024

Dataset Split

Train Set 92%
3210Images
Valid Set 5%
186Images
Test Set 3%
93Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 640x640

Augmentations

Outputs per training example: 3
Rotation: Between -15° and +15°
Shear: ±14° Horizontal, ±15° Vertical
Exposure: Between -15% and +15%
Blur: Up to 4.9px
Noise: Up to 1.96% of pixels