v2

2023-09-01 3:45pm

Generated on Sep 1, 2023

Dataset Split

Train Set 85%
11439Images
Valid Set 9%
1179Images
Test Set 7%
882Images

Preprocessing

Auto-Orient: Applied
Static Crop: 7-88% Horizontal Region, 6-90% Vertical Region
Resize: Stretch to 640x640
Tile: 3 rows x 3 columns

Augmentations

Outputs per training example: 3
Grayscale: Apply to 63% of images
Saturation: Between -33% and +33%