Dataset Versions

v1

2024-10-19 8:17am

Generated on Oct 19, 2024

Dataset Split

Train Set 97%
936Images
Valid Set 2%
24Images
Test Set 1%
6Images

Preprocessing

Resize: Stretch to 640x640
Auto-Adjust Contrast: Using Adaptive Equalization

Augmentations

Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Brightness: Between -20% and +20%
Blur: Up to 2.5px
Noise: Up to 1.6% of pixels