Dataset Versions
v5
2023-07-23 9:02pm
Generated on Jul 23, 2023
3717 Total Images
View All Images
Dataset Split
Train Set 97%
3609Images
Valid Set 1%
53Images
Test Set 1%
55Images
Preprocessing
Auto-Orient: Applied
Resize: Stretch to 640x640
Augmentations
Outputs per training example: 3
Rotation: Between -5° and +5°
Shear: ±10° Horizontal, ±7° Vertical
Grayscale: Apply to 23% of images
Hue: Between -28° and +28°
Saturation: Between -16% and +16%
Brightness: Between -22% and +22%
Exposure: Between -5% and +5%
Blur: Up to 1.25px
Noise: Up to 3% of pixels