Dataset Versions

v3

2024-01-31 3:46pm

Generated on Jan 31, 2024

Dataset Split

Train Set 89%
1629Images
Valid Set 7%
130Images
Test Set 4%
74Images

Preprocessing

Auto-Orient: Applied

Augmentations

Outputs per training example: 3
Flip: Horizontal
Crop: 0% Minimum Zoom, 33% Maximum Zoom
Rotation: Between -10° and +10°
Shear: ±18° Horizontal, ±15° Vertical
Saturation: Between -40% and +40%
Brightness: Between -33% and +33%
Exposure: Between -12% and +12%
Noise: Up to 3% of pixels