Matching datasets for Brain Computer Interfaces using Riemannian Geometry
| When |
May 23, 2019
from 11:00 to 12:00 |
|---|---|
| Where | R116 |
| Attendees |
Pedro Rodrigues |
| Add event to calendar |
vCal
iCal
|
In this talk, I will present a Transfer Learning approach for dealing with the statistical variability of EEG signals recorded on different sessions and/or from different subjects. This is a common problem faced by Brain-Computer Interfaces (BCI) and poses a challenge for systems that try to reuse data from previous recordings to avoid a calibration phase for new users or new sessions for the same user. I propose a method based on Procrustes analysis for matching the statistical distributions of two datasets using rigid geometrical transformations (translation, scaling and rotation) over the data points. The method uses symmetric positive definite matrices (SPD) as statistical features for describing the EEG signals, so the geometric operations on the data points respect the intrinsic geometry of the SPD manifold. Because of its geometry-aware nature, the method is named Riemannian Procrustes Analysis (RPA). The improvement in Transfer Learning via RPA is assessed by performing classification tasks on simulated data and on eight publicly available BCI datasets covering three experimental paradigms (243 subjects in total).