Dataset Versions

v3

2023-09-11 8:32am

Generated on Sep 11, 2023

Dataset Split

Train Set 89%
381Images
Valid Set 9%
38Images
Test Set 3%
11Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 640x640
Modify Classes: 0 remapped, 2 dropped

Augmentations

Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Rotation: Between -15° and +15°
Noise: Up to 5% of pixels