/BCI-2021-Riemannian-Geometry-workshop

Riemannian Geometry workshop at vBCI Meeting 2021

Primary LanguageJupyter Notebook

W13 Riemannian Geometry Methods and Tools for EEG preprocessing, analysis and classification

NOTE: if you participated to the workshop, please take 2min to give us your feedback to improve the material https://iyrp0g0k85t.typeform.com/to/g6Mcub4o

8th International BCI Meeting: 2021 Virtual BCI meeting, June 7 – 9, 2021

Workshop W13 Session 4- Thursday, June 9, 9:00am – 11:00am (Pacific Time) https://bcisociety.org/workshops/

Organizers:

  • Marco Congedo, Gipsa-lab/CNRS, Univ. Grenoble Alpes
  • Sylvain Chevallier, UVSQ, Université de Versailles Saint-Quentin-en-Yvelines
  • Louis Korczowski, Independent Scientist & Siopi.ai
  • Florian Yger, Université Paris-Dauphine
  • Pierre Clisson, Independent Scientist

Q&A

You can see all the questions and their answer at the workshop following this link. Moreover, you can directly comment the document if you want to ask more questions and add precisions.

Q&A: https://www.notion.so/Q-A-W13-Riemannian-Geometry-workshop-at-BCI-Meeting-2021-634290eefe0b47d59e0b7f68c9980ab6

News & Updates

Abstract

Riemannian Geometry (RG) is currently the object of growing interest within the BCI community. Machine learning methods based on RG have demonstrated robustness, accuracy and transfert learning capabilities for the classification of Motor Imagery, Event-Related Potential, Steady-State Visual Evoke Potentials, Sleep stages, as well as other mental states. This workshop will provide an overview of RG, emphasizing the characteristics that make RG compelling for BCI and its practical use for signal pre-processing, data analysis, mental state classification and regression. The workshop will be an opportunity for new users to solve real BCI problems (artifact removal, classification, transfert learning) with existing RG code resources and discuss the results.

Intended audience

BCI reasearchers/Neuroscientist working with EEG/MEG that are interested by Riemannian Geometry but haven’t used it already or who want a deeper understanding of the underlying properties (going beyong CSP)

Learning objectives

  1. Understanding Riemannian Geometry, its history of application in BCI
  2. Understanding the mathematical properties of Riemannian Geometry methods and the drawbacks
  3. Being able to find a use a specific method/toolbox using RG for a given need (preprocessing/analysis/classification/regression)

Timetable

Introduction

Part I: Talk and Q&A (50 min)

  • How Riemannian Geometry transformed BCI? (history, breakthroughs, example of applications) by Marco Congedo. Talk + Q&A.
  • Why Riemannian Geometry works so well? (properties, computational speed, etc.) by Florian Yger. Talk + Q&A.

Part II: Demonstrations & Discussions (50 min)

Because of the virtual format, we couldn't organize the coding sessions associated with the demonstration but all the code will be accessible after the workshop. Each demo will consist of a short introduction, code demo and Q&A.

  • "Automatical tag of your EEG artifacts with a Riemannian Potato and get better results using MNE+pyRiemann" by Louis Korczowski.
  • "Upgrading your standard classification pipeline with a simple RG trick using sklearn+pyRiemann" by Louis Korczowski.
  • "Evaluating BCI pipelines across datasets: the Mother of All BCI Benchmarks" by Sylvain Chevallier.
  • "Building a realtime realtime ERP speller using Riemannian Geometry and Timeflux" by Pierre Clisson.

Part III: Discussion on future challenges (10min)

Tools & Resources

Matlab

Python

R

REFERENCES

Barachant A, Bonnet S, Congedo M, Jutten C (2012) Multi-Class Brain Computer Interface Classification by Riemannian Geometry IEEE Transactions on Biomedical Engineering 59(4), 920-928.

Barachant A, Bonnet S, Congedo M, Jutten C (2013) Classification of covariance matrices using a Riemannian-based kernel for BCI applications Neurocomputing 112, 172-178.

Horev, I, Yger, F, Sugiyama, M (2016) Geometry-aware principal component analysis for symmetric positive definite matrices Machine LearningACML, 1-30

Mayaud L, Cabanilles S, Van Langhenhove A, Congedo M, Barachant A, Pouplin S, et al. (2016) Brain-computer interface for the communication of acute patients: a feasibility study and a randomized controlled trial comparing performance with healthy participants and a traditional assistive device Brain-Computer Interfaces, 3(4), 197-215.

Kalunga EK, Chevallier S, Djouani K, Monacelli E, Hamam Y (2016) Online SSVEP-based BCI using Riemannian Geometry Neurocomputing 191, 55-68

Yger, F., Berar, M., Lotte, F., (2017). Riemannian Approaches in Brain-Computer Interfaces: A Review. IEEE Transactions on Neural Systems and Rehabilitation Engineering 25, 1753–1762.

Congedo M, Barachant A, Bhatia R (2017) Riemannian Geometry for EEG-based Brain-Computer Interfaces; a Primer and a Review Brain-Computer Interfaces, 4(3), 155-174.

Congedo M, Barachant A, Kharati Koopaei E (2017) Fixed Point Algorithms for Estimating Power Means of Positive Definite Matrices IEEE Transactions on Signal Processing, 65(9), 2211-2220.

Bouchard F, Malick J, Congedo M (2018) Riemannian Optimization and Approximate Joint Diagonalization for Blind Source Separation IEEE Transactions on Signal Processing, 66 (8), 2041-2054.

Kalunga, EK, Chevallier, S, and Barthélemy, Q (2018) Transfer learning for SSVEP-based BCI using Riemannian similarities between users European Signal Processing Conference (EUSIPCO), 2018

Lotte F, Bougrain L, Cichocki A, Clerc M, Congedo M, Rakotomamonjy A, Yger F (2018) A Review of Classification Algorithms for EEG-based Brain-Computer Interfaces: A 10-year Update Journal of Neural Engineering, 15(3):031005.

Zanini P, Congedo M, Jutten C, Said S, Berthoumieu Y (2018) Transfer Learning: a Riemannian geometry framework with applications to Brain-Computer Interfaces IEEE Transactions on Biomedical Engineering, 65(5), 1107-1116.

Barthélemy Q, Mayaud Q, Ojeda D, Congedo M (2019) The Riemannian Potato Field: a tool for online Signal Quality Index of EEG IEEE Transactions on Neural Systems & Rehabilitation Engineering, 27 (2), 244-255 .

Bhatia R, Congedo M (2019) Procrustes problems in manifolds of positive definite matrices Linear Algebra and its Applications, 563, 440-445 .

Rodrigues PLC, Jutten C, Congedo M (2019) Riemannian Procrustes Analysis : Transfer Learning for Brain-Computer Interfaces IEEE Transactions on Biomedical Engineering, 66(8), 2390-2401.

Yger, F, Chevallier, S, Barthélemy, Q, and Sra, S (2020) Geodesically-convex optimization for averaging partially observed covariance matrices Asian Conference on Machine Learning

Khazem, S, Chevallier, S, Barthélemy, Q, Haroun, K, Noûs, C (2021) Minimizing Subject-dependent Calibration for BCI with Riemannian Transfer Learning IEEE EMBS NER

Chevallier S, Kalunga EK, Barthélemy Q, Monacelli E (2021) Review of Riemannian distances and divergences, applied to SSVEP-based BCI Neuroinformatics 19 (1), 93-106