Dataset Versions

v2

2023-09-01 3:40pm

Generated on Sep 1, 2023

Dataset Split

Train Set 79%
850Images
Valid Set 12%
131Images
Test Set 9%
98Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 640x640

Augmentations

Outputs per training example: 2
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Crop: 0% Minimum Zoom, 33% Maximum Zoom
Brightness: Between -14% and +14%
Blur: Up to 1.5px
Noise: Up to 18% of pixels
Cutout: 8 boxes with 20% size each