Dataset Versions
Versions
v1
2023-12-04 3:43pm
Generated on Dec 4, 2023
3240 Total Images
View All Images
Dataset Split
Train Set 88%
2835Images
Valid Set 8%
270Images
Test Set 4%
135Images
Preprocessing
Auto-Orient: Applied
Static Crop: 13-73% Horizontal Region, 23-78% Vertical Region
Resize: Fit (reflect edges) in 640x640
Tile: 5 rows x 1 columns
Augmentations
Outputs per training example: 3
Flip: Horizontal
Rotation: Between -9° and +9°
Shear: ±21° Horizontal, ±12° Vertical
Brightness: Between -39% and +39%
Exposure: Between -42% and +42%
Blur: Up to 2.5px
Noise: Up to 2% of pixels
Cutout: 11 boxes with 7% size each