QED-MultiClass Dataset

v4

2024-08-20 2:38pm

Generated on Aug 20, 2024

Dataset Split

Train Set 93%
1641Images
Valid Set 3%
61Images
Test Set 3%
61Images

Preprocessing

Static Crop: 23-77% Horizontal Region, 30-80% Vertical Region
Auto-Adjust Contrast: Using Histogram Equalization

Augmentations

Outputs per training example: 3
90° Rotate: Upside Down
Rotation: Between -8° and +8°

Similar Projects

See More