Dataset Versions
v2
2024-04-30 12:08am
Generated on Apr 29, 2024
1107 Total Images
View All Images
Dataset Split
Train Set 88%
972Images
Valid Set 8%
90Images
Test Set 4%
45Images
Preprocessing
Auto-Orient: Applied
Resize: Stretch to 640x640
Auto-Adjust Contrast: Using Adaptive Equalization
Tile: 3 rows x 3 columns
Augmentations
Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise
Rotation: Between -10° and +10°
Grayscale: Apply to 25% of images
Hue: Between -25° and +25°
Brightness: Between -15% and +15%
Blur: Up to 1.5px