System and method for multi-stage brain-computer interface training using neural networks
Abstract:
A system and method for a multi-stage brain-computer interface training using neural networks that reliably and predictably maps a user's thoughts to particular movements or actions in a computer-generated environment. The system comprises two stages: a pre-training stage, wherein specific exercises are generated on screen, and the brain activity is mapped to the exercises using a neural network as the user attempts to complete the exercises, and an in-use stage, wherein an initial mapping profile is loaded, brain activity is mapped to in-use interactions using a neural network, and those in-use mappings are compared to a library of stored mappings using a neural network to select a more accurate mapping for use in a given situation.
Information query
Patent Agency Ranking
0/0