Dataset Versions

v1

2023-05-06 7:20pm

Generated on May 6, 2023

Dataset Split

Train Set 95%
9648Images
Valid Set 3%
308Images
Test Set 2%
156Images

Preprocessing

Auto-Orient: Applied
Static Crop: 39-78% Horizontal Region, 25-75% Vertical Region
Resize: Stretch to 640x640
Tile: 2 rows x 2 columns

Augmentations

Outputs per training example: 3
Flip: Horizontal
90° Rotate: Clockwise, Counter-Clockwise
Shear: ±15° Horizontal, ±15° Vertical
Saturation: Between -25% and +25%
Brightness: Between -47% and +47%
Exposure: Between -35% and +35%
Blur: Up to 2.5px
Noise: Up to 5% of pixels

Similar Projects

See More