Embedding a Trained ML Model in OpenBCI GUI for Real-Time UDP Prediction Forwarding
Hey everyone! Could the OpenBCI GUI embed our trained machine learning model (using data collected by OpenBCI Ganglion) and feed real-time collected data into this model to obtain prediction results, and finally forward these prediction results in real time via the UDP protocol to an external device (an external device that is connected to the same WiFi network as the computer running the OpenBCI GUI) at its designated IP port? If not, what other methods can we use to achieve this? Thanks!!!
Comments
OpenBCI_GUI does support 'Networking Widget' output capabilities:
https://docs.openbci.com/Software/OpenBCISoftware/GUIWidgets/#networking
But there are no ML 'embedding' features. You might want to look at frameworks like NeuroPype or OpenViBE that use a flexible 'dataflow' diagram that you create between various signal processing 'boxes'.
https://openvibe.inria.fr/
https://www.neuropype.io/
You do not have to depend on the GUI to supply the sample stream. Instead you can call directly on Brainflow functions in the language of your choice to read the OpenBCI data. This is what the GUI is using under the covers.
https://brainflow.org/
https://brainflow.readthedocs.io/en/stable/Examples.html