Dataset Versions

v3

2024-09-23 6:34am

Generated on Sep 23, 2024

Dataset Split

Train Set 92%
183Images
Valid Set 4%
8Images
Test Set 4%
7Images

Preprocessing

Auto-Orient: Applied
Resize: Fit (white edges) in 640x640

Augmentations

Outputs per training example: 3
Rotation: Between -15° and +15°
Grayscale: Apply to 25% of images
Brightness: Between -40% and +40%
Exposure: Between -25% and +25%
Noise: Up to 2.52% of pixels
Cutout: 12 boxes with 6% size each