Dataset Versions
Versions
v1
Version 1.0
Generated on Dec 7, 2024
3073 Total Images
View All ImagesDataset Split
Train Set 87%
2688Images
Valid Set 8%
258Images
Test Set 4%
127Images
Preprocessing
Auto-Orient: Applied
Resize: Fit within 224x224
Auto-Adjust Contrast: Using Adaptive Equalization
Filter Null: Do not filter any null images.
Augmentations
Outputs per training example: 3
Crop: 0% Minimum Zoom, 20% Maximum Zoom
Blur: Up to 1px
Cutout: 3 boxes with 15% size each
Bounding Box: Shear: ±10° Horizontal, ±10° Vertical