CADDA: Class-wise Automatic Differentiable Data Augmentation for EEG Signals

TL;DR

This paper introduces CADDA, a differentiable method for automatic, class-wise data augmentation tailored for EEG signals, improving training efficiency and performance in neuroscience applications.

Contribution

It proposes a novel differentiable relaxation for class-wise data augmentation, enabling efficient search and application of augmentation policies for EEG data.

Findings

Outperforms gradient-free methods in class-wise augmentation tasks

Faster training with the new relaxation compared to previous gradient-based methods

Provides novel augmentation operations for sleep stage classification

Abstract

Data augmentation is a key element of deep learning pipelines, as it informs the network during training about transformations of the input data that keep the label unchanged. Manually finding adequate augmentation methods and parameters for a given pipeline is however rapidly cumbersome. In particular, while intuition can guide this decision for images, the design and choice of augmentation policies remains unclear for more complex types of data, such as neuroscience signals. Besides, class-dependent augmentation strategies have been surprisingly unexplored in the literature, although it is quite intuitive: changing the color of a car image does not change the object class to be predicted, but doing the same to the picture of an orange does. This paper investigates gradient-based automatic data augmentation algorithms amenable to class-wise policies with exponentially larger search spaces. Motivated by supervised learning applications using EEG signals for which good augmentation policies are mostly unknown, we propose a new differentiable relaxation of the problem. In the class-agnostic setting, results show that our new relaxation leads to optimal performance with faster training than competing gradient-based methods, while also outperforming gradient-free methods in the class-wise setting. This work proposes also novel differentiable augmentation operations relevant for sleep stage classification.