Dataset Versions
v2
2023-03-02 4:46pm
Generated on Mar 2, 2023
288 Total Images
View All Images
Dataset Split
Train Set 88%
252Images
Valid Set 8%
24Images
Test Set 4%
12Images
Preprocessing
Auto-Orient: Applied
Resize: Stretch to 640x640
Augmentations
Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Crop: 0% Minimum Zoom, 61% Maximum Zoom
Rotation: Between -21° and +21°
Shear: ±15° Horizontal, ±29° Vertical
Grayscale: Apply to 49% of images
Hue: Between -147° and +147°
Saturation: Between -63% and +63%
Blur: Up to 2.5px
Noise: Up to 6% of pixels
Cutout: 6 boxes with 11% size each