Dataset Versions

v1

2023-12-19 4:17pm

Generated on Dec 19, 2023

Dataset Split

Train Set 90%
2988Images
Valid Set 7%
237Images
Test Set 3%
95Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 1024x768
Auto-Adjust Contrast: Using Histogram Equalization

Augmentations

Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Blur: Up to 1px
Cutout: 6 boxes with 10% size each
Mosaic: Applied