Dataset Versions

v10

2024-01-19 9:49am

Generated on Jan 19, 2024

Dataset Split

Train Set 88%
4820Images
Valid Set 9%
476Images
Test Set 3%
151Images

Preprocessing

Resize: Stretch to 640x640
Auto-Adjust Contrast: Using Histogram Equalization
Modify Classes: 1 remapped, 1 dropped
Filter Null: Require all images to contain annotations.

Augmentations

Outputs per training example: 2
Hue: Between -20° and +20°
Saturation: Between -25% and +25%
Brightness: Between -15% and +15%