Card Detection Dataset

v4

2024-02-29 5:46pm

Generated on Mar 1, 2024

Dataset Split

Train Set 88%
375Images
Valid Set 8%
35Images
Test Set 4%
18Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 640x640

Augmentations

Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Grayscale: Apply to 19% of images
Bounding Box: Flip: Horizontal, Vertical
Bounding Box: Rotation: Between -18° and +18°