Dataset Versions

v3

2024-11-25 7:34pm

Generated on Nov 25, 2024

Dataset Split

Train Set 88%
970Images
Valid Set 8%
89Images
Test Set 4%
44Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 1280x720
Modify Classes: 0 remapped, 6 dropped
Filter Null: Require all images to contain annotations.

Augmentations

Outputs per training example: 3
Saturation: Between -3% and +3%
Brightness: Between -2% and +2%
Exposure: Between -1% and +1%