I am currently completing my PhD in mixed reality BCI. I am working on decoding upper limb motion from neural signals using embodied immersive experimental setups.

The video shared here is a 3 minute project pitch highlighting the experimental setup for user motor training combined with the neural network based decoder that trains on kinematic motion data and EEG brain signals. This has been achieved in a real-time closed loop BCI and is presented in a research software application framework (NeuroDojo).

In terms of architecture, we use Unity for the experimental paradigm and kinematic motion capture combined with Matlab/Simulink for EEG signal acquisition and a python based deep learning framework/decoder. Its all held together with tears and duck tape! But it works. Research papers will be linked below as they are published.