Dataset Versions

v6

2024-02-11 9:43pm

Generated on Feb 11, 2024

Dataset Split

Train Set 90%
2145Images
Valid Set 5%
108Images
Test Set 6%
132Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 640x640
Auto-Adjust Contrast: Using Histogram Equalization
Filter Null: Require all images to contain annotations.

Augmentations

Outputs per training example: 3
Saturation: Between -18% and +18%
Exposure: Between -14% and +14%