Dataset Versions

v1

First

Generated on Apr 15, 2024

Dataset Split

Train Set 91%
9762Images
Valid Set 5%
578Images
Test Set 4%
406Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 640x640
Filter Null: Require all images to contain annotations.

Augmentations

Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down