Dataset Versions

v16

last test

Generated on Jun 8, 2022

Dataset Split

Train Set 83%
612Images
Valid Set 12%
86Images
Test Set 6%
42Images

Preprocessing

Resize: Stretch to 2048x2048

Augmentations

Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Rotation: Between -45° and +45°
Shear: ±22° Horizontal, ±22° Vertical
Saturation: Between -99% and +99%
Exposure: Between -10% and +10%