Dataset Versions
v13
2023-01-04 9:47am
Generated on Jan 4, 2023
38520 Total Images
View All Images
Dataset Split
Train Set 84%
32184Images
Valid Set 15%
5832Images
Test Set 1%
504Images
Preprocessing
Isolate Objects: Applied
Resize: Fit within 320x320
Auto-Adjust Contrast: Using Histogram Equalization
Tile: 6 rows x 6 columns
Augmentations
Outputs per training example: 1
Grayscale: Apply to 50% of images
Saturation: Between -50% and +50%
Brightness: Between -33% and +33%
Exposure: Between -23% and +23%
Noise: Up to 13% of pixels