Dataset Versions

v6

2024-12-18 1:35am

Generated on Dec 17, 2024

Dataset Split

Train Set 91%
2148Images
Valid Set 6%
140Images
Test Set 3%
71Images

Preprocessing

Auto-Orient: Applied
Resize: Fit (white edges) in 1024x1024

Augmentations

Outputs per training example: 3
Flip: Horizontal
Rotation: Between -4° and +4°
Brightness: Between -6% and +6%
Noise: Up to 0.1% of pixels
Cutout: 3 boxes with 8% size each