2nd BCI-UC

The 2nd BCI-UC will be held on February 10th & 11th 2021 from 03:00 pm to 08:00 pm (CET).  Registration and abstract submission are open as of Dec 23rd, 2020:

Register & submit an abstract

For the 2nd BCI-UC, we solicit two types of submissions:

  • Presentations are applications for 20 minutes slots by a single presenter.
  • Mini-symposia are applications for 60 minutes slots by multiple presenters on a coherent topic, possibly including a panel discussion.

Voting on submitted abstracts begins on January 13th, 08:00 am (CET). Abstract submission remains open until January 25th at 11:59 (CET). However, we recommend submitting before the beginning of the voting period to ensure that every voter sees your submission. We will invite the authors of the top-voted abstracts to present at the un-conference. For further information, see the FAQ section.

Keynote Speakers


Directeur de Recherche, GIPSA-lab, CNRS, University Grenoble Alpes

Title: Recent Advances on Riemannian Transfer Learning

Abstract:  Recent advances on Riemannian transfer learning achieved in Grenoble will be presented. We will consider both the unsupervised and semi-supervised cross-subject and cross-session framework. The former is obtained by recentering both the source and target trial covariance matrices around their geometric mean in the manifold of positive definite matrices (PDMs). The latter by further stretching and rotating the target PDMs so as to match as close as possible the distribution of the source PDMs (Riemannian Procrustes Analysis). Then, we will present the dimensionality transcending method, which allows operating transfer learning when the number and/or location of the electrodes used in the target and source data do not match. Leveraging on these advances we will demonstrate the construction of a meta-database, i.e., merging many heterogeneous databases obtained by different experiments with different number and/or location of electrodes within the same brain-computer interface modality.

Biography: Marco Congedo obtained the Ph.D. degree in Biological Psychology with a minor in Statistics from the University of Tennessee, Knoxville, in 2003. From 2003 to 2006 he has been a post-doc fellow at the French National Institute for Research in Informatics and Control (INRIA) and at France Telecom R&D. From 2007 to 2020 Dr. he has been a Research Scientist at the “Centre National de la Recherche Scientifique” (CNRS) in the GIPSA Laboratory, Grenoble, France. Since 2020 he is a Research Director in the same institution. In 2013 he obtained the ‘Habilitation à Diriger des recherches’ from Grenoble Alpes University.

Dr. Congedo is interested in human electroencephalography (EEG), particularly in real-time EEG neuroimaging (neurofeedback and brain-computer interface) and mathematical tools useful for the analysis and classification of EEG data such as inverse solutions, blind source separation and Riemannian geometry. He has authored and co-authored over 150 scientific publications on these subjects.

Camille JEUNET

CNRS Research Scientist – Aquitaine Institute of Cognitive and Integrative Neuroscience (INCIA), Univ. Bordeaux / CNRS, France

Title: User-centered approach to improve BCI efficiency and usability: Stakes, Progress and Obstacles

Abstract:  EEG-based Mental-Task BCIs (MT-BCIs) are extremely promising, notably to restore or improve motor and cognitive performances, e.g., in stroke patients and athletes. Nonetheless, several scientific challenges still have to be taken up before these technologies are usable and actually used outside laboratories. We will focus on those challenges related to human learning. It is estimated that 10 to 30% of users are unable to control an MT-BCI. Understanding how we learn to self-regulate specific brain patterns to acheive such control, and which factors influence this learning, is thus essential in order to design acceptable and efficient MT-BCI training procedures. I will present recent progress made in the field, but also discuss with you the obstacles encountered including, inter alia, our poor understanding of the high within- and between-subject variability in terms of BCI performances, the usually small sample sizes or the limited reporting of the instructions provided during MT-BCI training procedures. I will argue that collaborative approaches and open science could help us overcoming these obstacles, and present an intiative in this direction.

Biography: Camille Jeunet received her PhD in Cognitive Sciences in 2016 at the University of Bordeaux, France. After a post-doctoral fellowship in Inria (Rennes, France) and EPFL (Geneva, Switzerland), she joined the University of Toulouse (France) as a tenured CNRS* Research Scientist. In 2021, she has rejoined the University of Bordeaux where she leads an interdisciplinary research bringing together computer sciences, psychology and neurosciences in order to better understand human learning mechanisms in BCIs, and to improve BCI user-training. She aims to identify and model the cognitive, psychological and neurophysiological factors that influence BCI performance and learning in order then to design innovative training procedures and feedback, adapted to each user. She is particularly interested in using EEG-based BCIs to improve cognitive and motor skills in athletes and stroke patients. Camille Jeunet has received 3 PhD awards as well as several national fundings for her research. Since 2017, she is board member of the French BCI association, CORTICO.


Contributed Talks & Mini-Symposia

... will be listed here after the voting period.