This repository contains the code to make the cyber player playing the mirror game in a group scenario. The group's members are simulated by different virtual trainers. Each virtual trainer is driven by an optimal control algorithm and has a markov chain model embedded in its core to mimic the kinematic features of a human player.
The paper related to this repository is: M. Lombardi, D. Liuzza, M. di Bernardo. "Dynamic input deep learning control of artificial avatars in a multi-agent joint motor task", submitted to Frontiers in Robotics and AI, 2021.