Viet-Anh on Software Logo

What is: RandAugment?

SourceRandAugment: Practical automated data augmentation with a reduced search space
Year2000
Data SourceCC BY-SA - https://paperswithcode.com

RandAugment is an automated data augmentation method. The search space for data augmentation has 2 interpretable hyperparameter NN and MM. NN is the number of augmentation transformations to apply sequentially, and MM is the magnitude for all the transformations. To reduce the parameter space but still maintain image diversity, learned policies and probabilities for applying each transformation are replaced with a parameter-free procedure of always selecting a transformation with uniform probability 1K\frac{1}{K}. Here KK is the number of transformation options. So given NN transformations for a training image, RandAugment may thus express KNKN potential policies.

Transformations applied include identity transformation, autoContrast, equalize, rotation, solarixation, colorjittering, posterizing, changing contrast, changing brightness, changing sharpness, shear-x, shear-y, translate-x, translate-y.