While interacting with virtual characters in a virtual environment or with a humanoid robot in the real world, several key enabling technologies are needed for the “suspension of disbelief” or feeling of presence (“being there”) and natural communication between the real user and the virtual or robotic equivalent. The aim of this project is the convergence of the most essential high-fidelity presence and interaction computer graphics and vision enabling technologies for improving human-computer and human-robot interaction in virtual as well as in augmented reality environments, involving novel virtual characters and humanoids. Both these human embodiments are very active research fields and recently their basic simulation technologies (face, body, behavior) started being applied from one domain (virtual) to the other (robotic).

On one hand, based on the extensive prior research experience of the fellow in computer graphics and real-time virtual character simulation technologies (realistic rendering, natural animation, and believable face & body simulation), a novel mathematical, unified, real-time simulation framework will be researched in this project as one key enabling technology in human-computer interaction. This framework will be based on Geometric Algebra aiming to avoid previous different, disjoint geometric graphics techniques by unifying these heterogeneous character simulation technologies to allow for their seamless abstraction and application in their humanoid robotic counterparts.

One the other hand, the ability of virtual characters or humanoids to visually perceive and understand articulated human motion (be that gestures, actions, expressions, etc) and to appropriately react or imitate this motion depending on the requirements of the specific task at hand is another key enabling technology in which the fellow will be trained explicitly by the host FORTH-Computational Vision and Robotics Laboratory in this project. So far the fellow has been utilizing extensively such vision-based techniques in his graphics and presence research career, but most often as “black-boxes” without having being formally introduced to their intrinsic characteristics. The host in this project will expose directly the fellow in the latest vision and robotics technologies and also empower him to directly implement them via the new unified framework in both robotic humanoids as well as ambient intelligent virtual characters. Via a specialized Career Development Plan, coupled with other vision, robotics and human physiology-related structured and complementary training, the host will provide the ideal theoretical as well as practical validation environment through which the fellow can reach new, unprecedented levels of professional maturity and significantly enhance and diversify his research career path.

Leave a Reply