Taking Control

Technical Papers

Taking Control

Tuesday, 11 August 3:45 PM - 5:35 PM | Los Angeles Convention Center, Room 152 Session Chair: Jehee Lee, Seoul National University


Hybrid Skeletal-Surface Motion Graphs for Character Animation From 4D Performance Capture

A novel hybrid representation for character animation from performance capture that combines skeletal control with surface motion graphs, extends the range and style of motion, and preserves the captured natural dynamics.

Peng Huang
University of Surrey

Margara Tejera
University of Surrey

John Collomosse
University of Surrey

Adrian Hilton
University of Surrey

Iterative Training of Dynamic Skills Inspired by Human Coaching Techniques

An intuitive, interactive framework for developing dynamic controllers, inspired by how humans learn dynamic motor skills through the progressive process of coaching and practice.

Sehoon Ha
Georgia Institute of Technology

C. Karen Liu
Georgia Institute of Technology

Dynamic Terrain Traversal Skills Using Reinforcement Learning

Applying reinforcement learning to develop controllers that allow physics-based characters to traverse terrains with gaps, steps, and walls using highly dynamic gaits.

Xue Bin Peng
The University of British Columbia

Glen Berseth
The University of British Columbia

Michiel van de Panne
The University of British Columbia

Online Control of Simulated Humanoids Using Particle Belief Propagation

This paper presents a novel, general-purpose model-predictive control (MPC) algorithm (Control Particle Belief Propagation) and demonstrates the method in balancing, recovery from both small and extreme disturbances, reaching, balancing on a ball, juggling a ball, and fully steerable locomotion in an environment with obstacles.

Perttu Hämäläinen
Aalto University

Joose Rajamäki
Aalto University

C. Karen Liu
Georgia Institute of Technology

Intuitive and Efficient Camera Control With the Toric Space

This paper introduces toric space, a novel and compact representation for intuitive and efficient virtual camera control. The proposed approach outperforms existing automated-viewpoint computation methods, provides an easy-to-use screen-space manipulation tool, and enables creation of complex camera motions such as long takes.

Christophe Lino
IRISA/INRIA Rennes Bretagne Atlantique

Marc Christie
University of Rennes1/IRISA