Dataset Versions

v2

2024-05-30 5:05am

Generated on May 30, 2024

Dataset Split

Train Set 92%
249Images
Valid Set 4%
10Images
Test Set 4%
11Images

Preprocessing

Resize: Fill (with center crop) in 640x640

Augmentations

Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Shear: ±14° Horizontal, ±14° Vertical
Grayscale: Apply to 10% of images
Hue: Between -20° and +20°
Saturation: Between -25% and +25%
Brightness: Between -10% and +10%
Cutout: 3 boxes with 12% size each
Bounding Box: Crop: 0% Minimum Zoom, 25% Maximum Zoom