Dataset Versions
v5
2024-12-14 1:57pm
Generated on Dec 14, 2024
1500 Total Images
View All Images
Dataset Split
Train Set 80%
1200Images
Valid Set 7%
100Images
Test Set 13%
200Images
Preprocessing
Auto-Orient: Applied
Static Crop: 31-67% Horizontal Region, 36-73% Vertical Region
Resize: Stretch to 1280x1280
Auto-Adjust Contrast: Using Adaptive Equalization
Tile: 10 rows x 10 columns
Augmentations
Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Shear: ±24° Horizontal, ±5° Vertical
Exposure: Between -39% and +39%