Dataset Versions

v7

2023-11-04 6:41pm

Generated on Nov 4, 2023

Dataset Split

Train Set 69%
87Images
Valid Set 16%
20Images
Test Set 16%
20Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 640x640

Augmentations

Outputs per training example: 3
Rotation: Between -35° and +35°
Grayscale: Apply to 12% of images
Hue: Between -102° and +102°
Saturation: Between -61% and +61%
Brightness: Between -60% and +60%
Exposure: Between -9% and +9%
Mosaic: Applied