Dataset Versions

v3

2024-04-15 11:56am

Generated on Apr 15, 2024

Dataset Split

Train Set 90%
3507Images
Valid Set 7%
261Images
Test Set 3%
135Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 640x640

Augmentations

Outputs per training example: 3
Flip: Horizontal
Shear: ±10° Horizontal, ±10° Vertical
Grayscale: Apply to 16% of images
Hue: Between -24° and +24°
Exposure: Between -12% and +12%