Dataset Versions

v9

2023-08-27 4:57am

Generated on Aug 27, 2023

Dataset Split

Train Set 92%
20184Images
Valid Set 7%
1612Images
Test Set 0%
60Images

Preprocessing

Auto-Orient: Applied
Resize: Fit within 1024x1024
Auto-Adjust Contrast: Using Adaptive Equalization
Tile: 2 rows x 2 columns

Augmentations

Outputs per training example: 3
Flip: Horizontal
Rotation: Between -30° and +30°
Hue: Between -3° and +3°
Saturation: Between -3% and +3%
Brightness: Between -10% and +10%
Exposure: Between -3% and +3%
Mosaic: Applied

Similar Projects

See More