Hello,
I am trying to run the motor-imagery with CSP scenarios (http://openvibe.inria.fr/motor-imagery- ... rn-filter/) with the Epoc headset.
After running motor-imagery-bci-1-acquisition, I receive an OV file for the CSP trainer and classifier trainer.
When I run motor-imagery-bci-2-train-CSP, the scenario runs forever. When i load the generic stream reader with the default demo file (bci-motor-imagery.ov), the scenario completes in 7 minutes so I know it should be working, but it fails when I select the acquired file from step 1.
Also, when I run motor-imagery-bci-3-classifier-training, i get the following error:
[ ERROR ] At time 0.023 sec <Box algorithm::CSP Spatial Filter> Bad matrix size - Expected 11 input channels and received 14 input channels
[WARNING] Box algorithm <CSP Spatial Filter> has been deactivated because process phase returned bad status
What am I doing wrong and how can i get this to work?
Any help would be greatly appreciated.
Motor Imagery scenarios on Emotiv Epoc
Re: Motor Imagery scenarios on Emotiv Epoc
Hi,
Maybe the last stimulation that should trigger the training is not in the recorded file. Try to change the train trigger to OVTK_StimulationId_ExperimentStop or OVTK_StimulationId_Train in csp spatial filter trainer box.
The other error should be solved once you computed the spatial filter on your data. the error is probably thrown because it uses the spatial filter trained on a different set with 11 input channels while you have 14.
Best regards,
Dieter
Maybe the last stimulation that should trigger the training is not in the recorded file. Try to change the train trigger to OVTK_StimulationId_ExperimentStop or OVTK_StimulationId_Train in csp spatial filter trainer box.
The other error should be solved once you computed the spatial filter on your data. the error is probably thrown because it uses the spatial filter trained on a different set with 11 input channels while you have 14.
Best regards,
Dieter
Re: Motor Imagery scenarios on Emotiv Epoc
Hello wuwu622,
the problem with CSP training is that the filter training can not converge (give a result) when one of the channels is a linear combination of the others. Since this is the case with the Emotiv EPOC device, you will have to use a channel selector to remove one of the acquired channels.
Once this is done you should be able to use the training scenario and subsequently the classifier training scenario.
Your error with motor-imagery-bci-3-classifier-training is due to the fact that you are trying to use the filter trained on a different combination of electrodes, since the motor-imagery-bci-2-train-CSP scenario never actually finished.
Hope this helps
Jozef
the problem with CSP training is that the filter training can not converge (give a result) when one of the channels is a linear combination of the others. Since this is the case with the Emotiv EPOC device, you will have to use a channel selector to remove one of the acquired channels.
Once this is done you should be able to use the training scenario and subsequently the classifier training scenario.
Your error with motor-imagery-bci-3-classifier-training is due to the fact that you are trying to use the filter trained on a different combination of electrodes, since the motor-imagery-bci-2-train-CSP scenario never actually finished.
Hope this helps
Jozef
Re: Motor Imagery scenarios on Emotiv Epoc
Hi all,
I am trying to use my EPOC to test the Motor-Imagery BCI example.(Not Motor-Imagery BCI CSP).
I can get the .ov data files from motor-imagery-bci-1-acquisition.xml. When I trying to run motor-imagery-bci-2-classifier-trainer.xml, it gave me the following error:
[ ERROR ] At time 0.086 sec <Box algorithm::Reference Channel> Channel not found [Nz]
[WARNING] Box algorithm <Reference Channel> has been deactivated because process phase returned bad status
Can someone tell me how to config the "Reference Channel" and "Channel Selector" boxes? What should I put there?
What's more, when I ran motor-imagery-bci-1-acquisition.xml, I always got:
[ INF ] At time 0.000 sec <Box algorithm::Generic stream writer> Compression flag not used yet, the file will be flagged uncompressed and stored as is
[ INF ] At time 0.000 sec <Box algorithm::Graz Motor Imagery BCI Stimulator> Lua script terminated
Does these matter?
Thank you so much.
I am trying to use my EPOC to test the Motor-Imagery BCI example.(Not Motor-Imagery BCI CSP).
I can get the .ov data files from motor-imagery-bci-1-acquisition.xml. When I trying to run motor-imagery-bci-2-classifier-trainer.xml, it gave me the following error:
[ ERROR ] At time 0.086 sec <Box algorithm::Reference Channel> Channel not found [Nz]
[WARNING] Box algorithm <Reference Channel> has been deactivated because process phase returned bad status
Can someone tell me how to config the "Reference Channel" and "Channel Selector" boxes? What should I put there?
What's more, when I ran motor-imagery-bci-1-acquisition.xml, I always got:
[ INF ] At time 0.000 sec <Box algorithm::Generic stream writer> Compression flag not used yet, the file will be flagged uncompressed and stored as is
[ INF ] At time 0.000 sec <Box algorithm::Graz Motor Imagery BCI Stimulator> Lua script terminated
Does these matter?
Thank you so much.
Re: Motor Imagery scenarios on Emotiv Epoc
No, the last two errors do not have any real effect. However, that Nz channel error, it is simple. You have to delete the reference channel box. There is no channel known as Nz is Emotiv.
Thanks
Thanks