Recent Advances in Virtual Avatar Technology
Sung-Hee Lee, KAIST
Creating Scene-Aware, User-Friendly Virtual Avatars
Lap-Fai (Craig) Yu, George Mason University
Abstract. In this talk, I will discuss recent advances of our group in creating virtual avatars that are aware of the semantics of real-world scenes and the perception of users, which lead to a number of interesting applications such as virtual pets and virtual assistants in mixed reality. Based on a 3D-reconstructed scene with objects recognized, a virtual avatar infers where to stand, what to do, and how to interact with the human user in a realistic and natural manner. I will also discuss inspirations on generating vivid voices for virtual avatars and devising realistic virtual coaches for training.
Fast and Flexible Multilegged Locomotion Using Learned Centroidal dynamics
Taesoo Kwon, Hanyang University
Abstract. We present a flexible and efficient approach for generating multilegged locomotion. Our model-predictive control (MPC) system efficiently generates terrain-adaptive motions, as computed using a three-level planning approach. This leverages two commonly-used simplified dynamics models, an inverted pendulum on a cart model (IPC) and a centroidal dynamics model (CDM). Taken together, these ensure efficient computation and physical fidelity of the resulting motion. The final full-body motion is generated using a novel momentum-mapped inverse kinematics solver and is responsive to external pushes by using CDM forward dynamics. For additional efficiency and robustness, we then learn a predictive model that then replaces two of the intermediate steps. We demonstrate the rich capabilities of the method by applying it to monopeds, bipeds, and quadrupeds, and showing that it can generate a very broad range of motions at interactive rates, including banked variable-terrain walking and running, hurdles, jumps, leaps, stepping stones, monkey bars, implicit quadruped gait transitions, moon gravity, push- responses, and more.
Breathing Life into Digital Characters
Seungbae Bang, University of Toronto
Abstract. Computer graphics industry is steadily growing up and the demand for a quality digital character is ever increasing. For the modeling and controlling of digital characters, many techniques are needed behind the scene. Despite the dazzling technical advances for this, most techniques currently adopted in the industry often require professional skills and intensive labor. I will introduce my work on providing users with better tools for modeling and controlling digital characters.