Data Augmentation

Apply random transformations to training patches to improve model generalization.

Data augmentation artificially expands the training dataset by applying random transformations to image patches during training.

Transformations applied

• Random rotations (90°, 180°, 270°)

• Horizontal and vertical flips

Benefits

• Helps the model generalize to different orientations

• Reduces overfitting on small datasets

• Improves robustness to image orientation

When to enable

• Most cases benefit from augmentation

• Especially helpful with smaller training datasets

When to disable

• autoStructN2V Stage 1: Disabled by default because augmentation can interfere with structural noise pattern detection

• If your images have a specific required orientation

Note: For autoStructN2V, augmentation is only available in Stage 2 (advanced options) after the noise pattern has been detected.