Homunculus (Direct in-mind physical skills upload/enhancing)

   System can teach people to enhance or learn new physical skills. 

   System allows pass teacher controlled training, verbal discussions, reference movement observation and "upload" skills directly in mind using direct visual and haptic impact on test subject.

  System teaching neural net subject's "writing" of target movement, which allows prediction in learning RT session how subject will moves. It allows system to reconfigure virtual world around subject forcing moves by new or optimal technique.

   Middle module proxies motion capture input and reconstruct virtual world coordinates is a main brain of system. It consists of RT proxy for mocap and AI that has human behavior model ("Homunculus") basing on prediction do some corrections to virtual world coordinate system and add delta ("DeJa Vu") between real and virtual world which forces subject move correctly or more efficiently.

   Test subject forced to correct his in-brain motor program and system decrease gap between worlds.

   Visual illusion system tricks users minds of their real movement. Using such illusions system going to force users move in according to calculated optimal path, which were get from optimization algorithm and input from neural network, trained by user's "writing", which gives opportunity to predict trajectory of feature user's movements. To do so there used this kind of architecture - motion capture data passes through proxy, and receives some modifications of real trajectories according to neural network and optimization algorithm input, and after that it passes to visualization system. System  modify users motion trajectory and world coordinates to achieve best results.

   In other world system uses motion sickness effect of VR systems to force users train target movements faster and more efficiently, changing their built-in motor program.