Hello everyone and happy new year,

I'm wondering about the possibility of feeding data streams from an
electroencephalography cap into the CLA to predict spatial/kinematic
information about the hands, arms, legs, etc. Similar to what this lab is
doing:

http://www.youtube.com/watch?v=9vVrCMUKtuA

I'm not an expert in the technical details of the CLA, but presumably one
would need to develop encoders for the EEG cap and the sensors for the body
parts you were looking to track.

Also, would there be any way of turning off the learning aspect of the CLA
(possibly by turning off the temporal pooler?) after a time so that one
would be able to generate this movement data solely by motor imagery, or am
I barking up the wrong tree as far as machine-learning algorithms go (which
would be a shame, considering the inherent neatness of analyzing the
neocortex with algorithms inspired by it)?

Thank you,
Marcos
_______________________________________________
nupic mailing list
[email protected]
http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org

Reply via email to