Dataset Versions

v3

2024-06-04 11:48am

Generated on Jun 4, 2024

Dataset Split

Train Set 88%
1701Images
Valid Set 8%
159Images
Test Set 4%
81Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 640x640

Augmentations

Outputs per training example: 3
Crop: 0% Minimum Zoom, 29% Maximum Zoom
Shear: ±10° Horizontal, ±10° Vertical
Hue: Between -15° and +15°
Brightness: Between -18% and +18%
Blur: Up to 1.9px
Noise: Up to 1.25% of pixels