I am a Cornell PhD with interest in BCI+machine learning=mind reading.
Emotiv has portable EEG on market that can send EEG data wirelessly to machine, so I think it not hard to add a portable switch to record mind label simultaneously. For example, when feeling pleasant (or annoyed, desired to smoke, etc.), the user can immediately switch to 1, and switch to 0 if not feeling so. Or, more generally, the switch can have more states:
0=none, 1=a little, … 10=extremely. The switch’s state, representing mind label, is recorded simultaneously with EEG.
Then, machine learning can be used to decode such mind states.
Compared with lab experiment, it is less costly, more flexible, more convenient to enrich the training data, and easier to detect some emotions like burst of anger. In addition, the training sample collection gives users a convenient way for self-reflection.
As a result, I consider the daily-life experiment framework valuable.
Could anyone forward this idea to the hardware engineers, so they could design such simultaneous switches? Perhaps you could also tell the software engineers to link the state of switches to software, and match the time points of switch states and EEG.
If successful, the new experiment framework can perhaps become popular, and enrich Emotiv.
Thank you.
