Dataset Versions

v1

2023-05-06 6:22pm

Generated on May 6, 2023

Dataset Split

Train Set 93%
5877Images
Valid Set 3%
190Images
Test Set 4%
238Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 640x640

Augmentations

Outputs per training example: 3
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Hue: Between -100° and +100°
Saturation: Between -70% and +70%
Brightness: Between -33% and +33%
Noise: Up to 4% of pixels
Cutout: 9 boxes with 10% size each