Dataset Versions

v3

2024-09-01 9:34am

Generated on Sep 1, 2024

Dataset Split

Train Set 92%
2583Images
Valid Set 6%
164Images
Test Set 2%
50Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 640x640

Augmentations

Outputs per training example: 3
Crop: 0% Minimum Zoom, 25% Maximum Zoom
Shear: ±5° Horizontal, ±8° Vertical
Grayscale: Apply to 5% of images
Brightness: Between -10% and +10%
Exposure: Between -7% and +7%
Blur: Up to 0.5px
Noise: Up to 1.05% of pixels