Dataset Versions
v5
Strong
Generated on Mar 22, 2024
15424 Total Images
View All Images
Dataset Split
Train Set 87%
13440Images
Valid Set 7%
1088Images
Test Set 6%
896Images
Preprocessing
Auto-Orient: Applied
Tile: 8 rows x 8 columns
Augmentations
Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise
Crop: 0% Minimum Zoom, 27% Maximum Zoom
Rotation: Between -13° and +13°
Hue: Between -23° and +23°
Brightness: Between -16% and +0%
Blur: Up to 4.9px
Noise: Up to 1.96% of pixels