Dataset Versions

v3

2023-08-12 10:38pm

Generated on Aug 12, 2023

Dataset Split

Train Set 88%
138Images
Valid Set %
0Images
Test Set 12%
19Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 224x224

Augmentations

Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Upside Down
Rotation: Between -5° and +5°
Hue: Between -25° and +25°
Saturation: Between -5% and +5%
Brightness: Between -36% and +36%
Exposure: Between -25% and +25%
Noise: Up to 3% of pixels