New tutorial: Recording Event Related Potentials (ERPs)

About BCI and box tutorial/demo scenarios bundled with OpenViBE.
Post Reply
jtlindgren
Posts: 778
Joined: Tue Dec 04, 2012 3:53 pm
Location: INRIA Rennes, FRANCE

New tutorial: Recording Event Related Potentials (ERPs)

Post by jtlindgren »

Hi all,

I've posted a little tutorial how to record Event Related Potentials (ERPs) easily with OpenViBE. The provided scenario is simple to customize to collect other kinds of event-driven data as well.

You can find the stuff here!

Cheers,
Jussi

hackr
Posts: 12
Joined: Fri Aug 25, 2017 1:41 am

Re: New tutorial: Recording Event Related Potentials (ERPs)

Post by hackr »

@jtlindgren Thank you for this great tutorial! It's the most useful I've found thusfar.

2 questions:

1. Was there an example Lua script to go along with it?
2. I want to do what you did in the tutorial, but proceed to training, then classification, then execute commands/scripts based on those classifications rather than recording. Can you recommend any other tutorials or learning materials to use in combination with your tutorial to achieve these goals?

Thanks!

Update: I've found an ancient tutorial video and the classification-0-example.xml scenario to help me, but I'm still open to suggestions. Man, there is a lot to learn!

jtlindgren
Posts: 778
Joined: Tue Dec 04, 2012 3:53 pm
Location: INRIA Rennes, FRANCE

Re: New tutorial: Recording Event Related Potentials (ERPs)

Post by jtlindgren »

Hello Hackr,

the ERP Tutorial scenarios along with the .lua script are included in the OpenViBE 2.0 BETA if you dare to try that. :)

The P300 magic card scenarios are also using cue display, and they have some classification stuff inbuilt. We haven't used that scenario for a while so I'm not sure how well it works. In any case the basic idea to turn it into a classifier is the following,

1) You make a data collection scenario that displays images and records the signal and the timeline
3) You make another scenario that extracts features from the signal during relevant times and then trains a classifier with the feature vectors.
3) You use the previously trained classifier online to classify new data with features extracted in the same way. How you show the images in this case is not that clear, it might depend on what you want to do.

The ERP tutorial does the part 1, the parts 2 and 3 you'd have to think how to do them (look at motor imagery scenarios, read literature, etc). Its an active research subject.


Happy haxoring,
Jussi

Post Reply