Dataset Versions

v1

2022-10-17 4:25pm

Generated on Oct 17, 2022

Dataset Split

Train Set 88%
2100Images
Valid Set 8%
200Images
Test Set 4%
100Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 256x256

Augmentations

Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Grayscale: Apply to 25% of images
Cutout: 3 boxes with 10% size each