Dataset Versions

v8

v5

Generated on Sep 18, 2023

Dataset Split

Train Set 88%
2481Images
Valid Set 8%
236Images
Test Set 4%
118Images

Preprocessing

Auto-Orient: Applied
Resize: Fit (black edges) in 640x640

Augmentations

Outputs per training example: 3
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Rotation: Between -45° and +45°
Shear: ±10° Horizontal, ±10° Vertical
Brightness: Between -20% and +20%
Exposure: Between -27% and +27%
Blur: Up to 1px
Noise: Up to 5% of pixels