174. All The Brain Chip Implant Benefits & None of The Risks

Engineering & simulationEEG Net25% error reduction after 7 trials

The team developed a proprietary deep learning model called EEG Net to decode non-invasive brain signals, which reduced tracking error by 25% after just seven user trials compared to state-of-the-art baselines.

“they were basically on a relative basis able to reduce error by about 25% in a single user setting after 7 trials. So this means, um, if I were to sit there and use their EEG net model … the amount of error would be reduced by about 25% after seven, just 7 times of me trying the game.”

Engineering & simulationPointNet

The team applied the open-source PointNet model, typically used for 3D visualization, to correlate 3D brain activity locations with cursor movement vectors, improving the translation of neural signals to computer control.

“they also pointed, like you're saying, to this open source model called Pointnet, which is basically used to help visualize differences between different locations of items in 3D space. … use point net to correlate between these … create the basically the vector between those 2 points in 3D space and then use that to translate, what is the brain trying to do to control the computer?”

Engineering & simulationTransfer Learning4 trials vs 7 trials

By using transfer learning, the AI model applied knowledge from one user's training data to a new user, allowing the new user to achieve similar performance in only four trials instead of the seven required for a fresh model.

“they were able to achieve similar performance is they did in the single user setting after 7 trials, when they did transfer learning, they were able to get similar performance within only 4 trials. So basically this means like for each incremental user … it gets better and better at learning faster, and people achieve better and better performance within a shorter period of time, with less training rounds.”