Dataset Versions
v2
2024-06-26 6:45am
Generated on Jun 26, 2024
7984 Total Images
View All Images
Dataset Split
Train Set 89%
7122Images
Valid Set 7%
541Images
Test Set 4%
321Images
Preprocessing
Auto-Orient: Applied
Tile: 4 rows x 4 columns
Filter Null: Require at least 50% of images to contain annotations.
Augmentations
Outputs per training example: 4
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise
Saturation: Between -25% and +25%
Cutout: 4 boxes with 10% size each