Home Uncategorized Artificial Neural Network to Control Two Robotic Hands

Artificial Neural Network to Control Two Robotic Hands

64
0

This video shows application of trained artificial neural network algorithms to control two robotic hands using recorded motor imaginary signals from human brain. The team used 16 channels on motor and sensory cortex region for neuroscience study and 3 channels (C3,C4 and FCz) for real time Brain-Machine Interface experiment. Sixteen features were extracted from Alpha and Beta waves of the EEG signals to achieve overall 95% accuracy using Scaled Conjugate Gradient, Levenberg-Marquardt, and Bayesian Regularization methods. For more info check out this website: http://webs.wichita.edu/?u=jd_bme&p=/research/