Dataset Versions
v8
2024-08-28 2:08am
Generated on Aug 28, 2024
1184 Total Images
View All ImagesDataset Split
Train Set 78%
925Images
Valid Set 11%
131Images
Test Set 11%
128Images
Preprocessing
Auto-Orient: Applied
Resize: Stretch to 640x640
Filter Null: Require at least 90% of images to contain annotations.
Augmentations
Outputs per training example: 3
Crop: 0% Minimum Zoom, 30% Maximum Zoom
Shear: ±5° Horizontal, ±5° Vertical
Blur: Up to 2px
Noise: Up to 1% of pixels