/idport-api

WIP API of IDport.

Primary LanguagePython

What is this?

This is the work-in-progress (WIP) repository for the brain-computer interface API of Senzing.

A brain-computer interface (BCI) translates brain activity into a control signal. It thus holds the promise to control things with your thoughts alone! Although we made quite some progress, we can't really read your minds. But some signals we can detect reliably.

To explain the use of this API, we first need to outline several components that need to work together for the BCI to work.

The end-user application

Ultimately, we want to control something with our thoughts, for example a game. Somehow, this game has to be made aware of the state of our brain. This can be done by asking the IDport server to analyze the user's current brain activity.

The pattern recognizer

The analysis of brain activity is done by a series of automatic detectors, that process incoming EEG signals. These detectors are trained on lots of example recordings using advanced machine learning methods. By training we mean that the algorithms automatically learn to recognize important patterns. Distinguishing for Senzing's IDport is that we work with spontaneous brain activity, and that recognize brain activity without callibration.

Headset

We haven't explained where these electrical brain-signals signals come from. Theses EEG signals can be picked up by sensors placed on your head that measure the tiny potentials that your brain constantly emits. Research-grade EEG sets are quite expensive, but cheaper alternatives aimed at consumers do exist. But generally the quality of the signals is a bit worse than the research sets.

The most prominent consumers headsets that are the Emotiv EPOC, the NeuroSky MindWave and the InteraXon Muse.

Fusing the component

The application, the pattern recognizer and the EEG headset have to work together for the brain-computer interface to work. We have chosen to build a central web service that can be asked to analyze the users brain. That is all there exists from the perspective of the application.

For the EEG headset a small driver has to be written that writes incoming samples to the same web service. As before, this driver only communicates with the web service.

Finally, behind the scenes, this web service communicates with our pattern detectors.

The IDPort API facilitates communication between these components.