Dataset Versions

v12

2023-05-24 11:38am

Generated on May 24, 2023

Dataset Split

Train Set 89%
516Images
Valid Set 6%
36Images
Test Set 4%
26Images

Preprocessing

Auto-Orient: Applied
Static Crop: 15-83% Horizontal Region, 0-97% Vertical Region
Resize: Stretch to 608x608
Modify Classes: 2 remapped, 2 dropped

Augmentations

Outputs per training example: 3
90° Rotate: Clockwise, Counter-Clockwise
Shear: ±15° Horizontal, ±15° Vertical
Saturation: Between -25% and +25%
Bounding Box: Shear: ±15° Horizontal, ±15° Vertical