Dataset Versions
v3
2022-07-05 1:50pm
Generated on Jul 5, 2022
4532 Total Images
View All Images
Dataset Split
Train Set 85%
3864Images
Valid Set 6%
292Images
Test Set 8%
376Images
Preprocessing
Auto-Orient: Applied
Resize: Fit within 416x416
Tile: 2 rows x 2 columns
Modify Classes: 0 remapped, 1 dropped
Augmentations
Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Crop: 0% Minimum Zoom, 25% Maximum Zoom
Rotation: Between -17° and +17°
Shear: ±15° Horizontal, ±17° Vertical
Grayscale: Apply to 2% of images
Mosaic: Applied