Search found 160 matches

by ddvlamin
Tue Apr 02, 2013 1:57 pm
Forum: Scenarios and BCI design
Topic: P300 Speller, Problem with Online Part
Replies: 1
Views: 4510

Re: P300 Speller, Problem with Online Part

Did you change anything to the scenarios? It seems you get different number of features when doing the online experiment (compared to the training data), maybe you changed the epoch length, the decimation factor or you have a different number of channels that are coming from the acquisition client?
by ddvlamin
Tue Apr 02, 2013 1:55 pm
Forum: Scenarios and BCI design
Topic: P300 Online Speller
Replies: 1
Views: 4992

Re: P300 Online Speller

You should make sure that the number of repetitions in your P300 stimulator corresponds to the number of repetitions in your voting classifier if no averaging of the epochs is used. If you average all the repetitions, then you only need to vote once to get a result. If you have 12 repetitions and yo...
by ddvlamin
Tue Apr 02, 2013 1:32 pm
Forum: Boxes
Topic: csp
Replies: 4
Views: 11185

Re: csp

Probably this comes a bit late. Anyway, I wrote that code a while ago. There are indeed some things that have been done differently and could be improved/changed. Each epoch of each class is first centred, then the covariance is computed and then this is trace normalized. However in OpenViBE I thoug...
by ddvlamin
Thu Nov 08, 2012 8:41 am
Forum: Designer
Topic: --no-gui --play versus --play
Replies: 0
Views: 4902

--no-gui --play versus --play

Hi, I encountered a weird problem that is hard to reproduce. When I play a certain scenario (to train for example the classifier for the P300) with the --play option at the command-line then everything works fine all the time. However, when I want to run it with --no-gui --play, then sometimes the c...
by ddvlamin
Thu Nov 08, 2012 8:26 am
Forum: Boxes
Topic: P300 Magic card visualisition
Replies: 3
Views: 10947

Re: P300 Magic card visualisition

Hi, I would need to take a close look too as I don't know the c code behind that visualiser, but here is where you might start: you could try to change the p300-magic-card.ui file in the share/openvibe-plugins/simple-visualisation directory You could use Glade for that or do it manually in the file....
by ddvlamin
Tue Nov 06, 2012 1:07 pm
Forum: Boxes
Topic: plotting with the python box
Replies: 3
Views: 7536

Re: plotting with the python box

Hi, I still have this problem when closing openvibe after running a scenario with the python box. I do not only have it on my linux machine but also on the windows 7 machines. Anyone that can give me a hint where this problem comes from? If it worked for you, what python version did you use and how ...
by ddvlamin
Thu Sep 27, 2012 12:37 pm
Forum: Designer
Topic: question about time
Replies: 8
Views: 15099

Re: question about time

We got our algorithm to work properly without depending on the timing of certain incoming stimuli, so we avoided the problem You should look at the Generic Stream Reader box source code, and maybe the writer source code as well. This will help you understand how it works. Ok, will do. Thanks, Dieter
by ddvlamin
Thu Sep 27, 2012 12:28 pm
Forum: Box and application development
Topic: boost library
Replies: 0
Views: 5875

boost library

Hi, I wanted to start a thread in a box (just to experiment with some things) and of course wanted to use the boost library for this. I had to change the cmake file that finds the thread library a bit so that it chose the right boost dependency that comes with openvibe and not the one that comes wit...
by ddvlamin
Mon Sep 03, 2012 8:12 am
Forum: Discussion about BCI and related topics
Topic: BCI results display
Replies: 8
Views: 10670

Re: BCI results display

Hi,

I think for this you just need to add the Graz visualisation box to the scenario

Dieter
by ddvlamin
Mon Sep 03, 2012 8:00 am
Forum: Designer
Topic: question about time
Replies: 8
Views: 15099

Re: question about time

I am surprised how much these timing concerns raise questions, dicussions and explanations. I'm sorry about these questions, but I guess if someone writes a new box and notices some things while debugging that are not well understood, one will try to understand what is in the black box. Sometimes i...
by ddvlamin
Tue Aug 21, 2012 9:05 am
Forum: Designer
Topic: question about time
Replies: 8
Views: 15099

Re: question about time

Dear Yann, I agree and understand what you say (although not in depth as I still have some questions ;) sorry ), but in my opinion it still does not explain the 300 ms of delay (to me 300ms is a century in computer time). The two scenarios are basically the same, one with a file reader and an epochi...
by ddvlamin
Fri Aug 17, 2012 2:03 pm
Forum: Designer
Topic: question about time
Replies: 8
Views: 15099

Re: question about time

We saw that you added the 2 debug messages, the first one being already in the original source code. I hope you don't mean with "we saw" that I accidently commited any like that? In the last debug message the difference between the chunk start time and end time is not 1 second. Are you printing the...
by ddvlamin
Wed Aug 15, 2012 11:22 am
Forum: Designer
Topic: question about time
Replies: 8
Views: 15099

question about time

Hi, We are testing some new algorithms on P300 data that we previously recorded. However we had some weird problems and consequently we set up a simple scenario with a generic stream reader (that reads the P300 file) and an stimulation based epoch box, we added some debug output to the epoching box ...
by ddvlamin
Wed Aug 15, 2012 10:45 am
Forum: Scenarios and BCI design
Topic: Motor imagery
Replies: 1
Views: 4550

Re: Motor imagery

There should be motor-imagery scenarios in the share/openvibe-scenarios/bci directory
by ddvlamin
Wed Aug 15, 2012 10:41 am
Forum: Designer
Topic: P-300 Speller: bad results
Replies: 1
Views: 4576

Re: P-300 Speller: bad results

I haven't taken a look at your scenarios, but I assume you started from the ones that came with openvibe. Given that you use the Emotiv set, which generally has lower SNR, I would go for at least 20 repetitions, where you might want to use averaging instead of single trial analysis and voting. If yo...