Dataset Versions

v2

2024-08-22 3:37pm

Generated on Aug 22, 2024

Dataset Split

Train Set 97%
528Images
Valid Set 2%
12Images
Test Set 1%
4Images

Preprocessing

Auto-Orient: Applied
Static Crop: 25-75% Horizontal Region, 25-75% Vertical Region
Resize: Fill (with center crop) in 640x640
Auto-Adjust Contrast: Using Histogram Equalization
Tile: 2 rows x 2 columns

Augmentations

Outputs per training example: 3
Rotation: Between -15° and +15°
Shear: ±13° Horizontal, ±12° Vertical
Noise: Up to 1.85% of pixels