Dataset Versions
v2
2023-11-03 2:37pm
Generated on Nov 3, 2023
1491 Total Images
View All Images
Dataset Split
Train Set 88%
1305Images
Valid Set 6%
93Images
Test Set 6%
93Images
Preprocessing
Auto-Orient: Applied
Resize: Stretch to 120x120
Augmentations
Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Crop: 0% Minimum Zoom, 28% Maximum Zoom
Rotation: Between -38° and +38°
Grayscale: Apply to 100% of images
Hue: Between -177° and +177°
Saturation: Between -92% and +92%
Brightness: Between -33% and +33%
Blur: Up to 4.75px
Noise: Up to 15% of pixels
Cutout: 13 boxes with 14% size each