Dataset Versions
v2
2023-04-03 9:23am
Generated on Apr 3, 2023
28000 Total Images
View All Images
Dataset Split
Train Set 57%
16000Images
Valid Set 27%
7600Images
Test Set 16%
4400Images
Preprocessing
Auto-Orient: Applied
Resize: Stretch to 640x640
Tile: 20 rows x 20 columns
Augmentations
Outputs per training example: 1
Flip: Horizontal
Crop: 0% Minimum Zoom, 56% Maximum Zoom
Rotation: Between -27° and +27°
Shear: ±6° Horizontal, ±7° Vertical
Brightness: Between -12% and +12%
Exposure: Between -15% and +15%