Dataset Versions
Versions
v3
DatasetV1
Generated on Mar 10, 2024
658 Total Images
View All Images
Dataset Split
Train Set 93%
612Images
Valid Set 6%
38Images
Test Set 1%
8Images
Preprocessing
Auto-Orient: Applied
Resize: Stretch to 640x640
Augmentations
Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Crop: 0% Minimum Zoom, 50% Maximum Zoom
Rotation: Between -25° and +25°
Shear: ±10° Horizontal, ±20° Vertical
Grayscale: Apply to 19% of images
Hue: Between -90° and +90°
Saturation: Between -60% and +60%
Brightness: Between -50% and +50%
Exposure: Between -30% and +30%
Blur: Up to 2px
Noise: Up to 5.01% of pixels
Cutout: 15 boxes with 7% size each
Mosaic: Applied
Bounding Box: Flip: Horizontal, Vertical
Bounding Box: 90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Bounding Box: Crop: 0% Minimum Zoom, 15% Maximum Zoom
Bounding Box: Rotation: Between -10° and +10°
Bounding Box: Shear: ±5° Horizontal, ±5° Vertical
Bounding Box: Brightness: Between -15% and +15%
Bounding Box: Exposure: Between -10% and +10%
Bounding Box: Blur: Up to 0.5px
Bounding Box: Noise: Up to 0.14% of pixels