Data Augmentation
Recent work has shown that data augmentation has the potential to significantly improve the generalization of deep learning models. Recently, automated augmentation strategies have led to state-of-the-art results in computer vision tasks. However, when it comes to adopting this to a large scale, the separation of the search phase increases the training complexity leading to an increase in computational cost.
RandAugment significantly reduces the search space which allows the model to train on the target task with no need for a separate proxy task.
Parameters
All you have to do is set 2 parameters.
1. The number of augmentation transformations to apply sequentially.
2. The magnitude of all the transformations.
Reference
RandAugment: Practical automated data augmentation
with a reduced search space
