Dataset Versions

v3

2024-10-29 5:20am

Generated on Oct 29, 2024

Dataset Split

Train Set 86%
834Images
Valid Set 9%
88Images
Test Set 5%
44Images

Preprocessing

Auto-Adjust Contrast: Using Histogram Equalization

Augmentations

Outputs per training example: 3
Flip: Horizontal, Vertical
Brightness: Between -5% and +5%
Exposure: Between -5% and +5%
Noise: Up to 0.5% of pixels