Dataset Versions
v6
2024-10-23 5:46am
Generated on Oct 23, 2024
720 Total Images
View All Images
Dataset Split
Train Set 88%
630Images
Valid Set 8%
60Images
Test Set 4%
30Images
Preprocessing
Auto-Orient: Applied
Resize: Stretch to 640x640
Augmentations
Outputs per training example: 3
Crop: 5% Minimum Zoom, 27% Maximum Zoom
Grayscale: Apply to 13% of images
Hue: Between -20° and +20°
Saturation: Between -31% and +31%
Brightness: Between -24% and +24%
Exposure: Between -12% and +12%
Blur: Up to 1.1px
Noise: Up to 1.6% of pixels