Data Augmentation for Brain-Computer Interfaces: Analysis on Event-Related Potentials Data
Computer Vision and Pattern Recognition
2018-01-10 v1 Neurons and Cognition
Abstract
On image data, data augmentation is becoming less relevant due to the large amount of available training data and regularization techniques. Common approaches are moving windows (cropping), scaling, affine distortions, random noise, and elastic deformations. For electroencephalographic data, the lack of sufficient training data is still a major issue. We suggest and evaluate different approaches to generate augmented data using temporal and spatial/rotational distortions. Our results on the perception of rare stimuli (P300 data) and movement prediction (MRCP data) show that these approaches are feasible and can significantly increase the performance of signal processing chains for brain-computer interfaces by 1% to 6%.
Cite
@article{arxiv.1801.02730,
title = {Data Augmentation for Brain-Computer Interfaces: Analysis on Event-Related Potentials Data},
author = {Mario Michael Krell and Anett Seeland and Su Kyoung Kim},
journal= {arXiv preprint arXiv:1801.02730},
year = {2018}
}