Dataset Versions

v7

finalvalidation

Generated on Mar 21, 2023

Dataset Split

Train Set %
0Images
Valid Set 100%
255Images
Test Set %
0Images

Preprocessing

Auto-Adjust Contrast: Using Histogram Equalization
Filter Null: Require at least 98% of images to contain annotations.

Augmentations

Outputs per training example: 1
Flip: Horizontal
Grayscale: Apply to 25% of images
Exposure: Between -10% and +10%