NEURO-LOOP
Still being built. Real-time BCI for human-in-the-loop RL.
NEURO-LOOP is a reinforcement learning framework that integrates implicit neural feedback into robotic and artificial agent policy learning. In this iteration, we explore using a hybrid fNIRS-EEG system to train the agent in a real-time loop. Further, we classify multiple brain states and signals to further improve the agent's performance.
Previous work used passive neural activity as a supervisory signal so agents could adapt to a user’s expectations as they watch the agent. Current work uses a hybrid fNIRS-EEG system to classify performance, preference, error, and attentional states, and feed that into a robotic agent’s policy.
We plan to release a dataset of fNIRS and EEG from the human-robot interaction, with the robot’s learning statistics alongside the brain data, then run the framework in real time to see how it performs.
Currently under development.