Dataset Versions

v3

2024-07-28 7:24pm

Generated on Jul 28, 2024

Dataset Split

Train Set 87%
4784Images
Valid Set 8%
458Images
Test Set 4%
240Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 640x640

Augmentations

Outputs per training example: 4
Rotation: Between -20° and +20°
Grayscale: Apply to 25% of images
Saturation: Between -25% and +25%
Brightness: Between -25% and +25%
Exposure: Between -25% and +25%
Blur: Up to 10px
Noise: Up to 5.01% of pixels