Dataset Versions

v1

2024-02-26 11:04am

Generated on Feb 26, 2024

Dataset Split

Train Set 85%
13962Images
Valid Set 13%
2151Images
Test Set 2%
358Images

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
Brightness: Between -20% and +20%
Noise: Up to 1.21% of pixels