Dataset Versions

v59

2024-03-21 10:53pm

Generated on Mar 21, 2024

Dataset Split

Train Set 86%
6684Images
Valid Set 10%
778Images
Test Set 4%
336Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 640x640
Auto-Adjust Contrast: Using Histogram Equalization

Augmentations

Outputs per training example: 3
Rotation: Between -4° and +4°
Grayscale: Apply to 100% of images
Saturation: Between -30% and +30%
Brightness: Between 0% and +23%
Exposure: Between -10% and +10%