Dataset Versions

v5

tile-Donly

Generated on May 2, 2024

Dataset Split

Train Set 70%
763Images
Valid Set 16%
176Images
Test Set 13%
145Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 640x640
Tile: 4 rows x 4 columns
Filter Null: Require at least 40% of images to contain annotations.

Augmentations

Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Grayscale: Apply to 15% of images
Saturation: Between -15% and +15%
Brightness: Between -15% and +15%

Similar Projects

See More