Performance
Animation from Low-dimensional Control Signals
Real-time
animation and control of three-dimensional human motions using low-cost and
non-intrusive devices
|
|
Project
description
The ability to accurately reconstruct a user's motion in real time would
allow the intuitive control of characters in computer games, the control of
avatars for virtual reality or electronically mediated communication, and the
rapid prototyping of character animations. This project introduces an approach
to performance animation that employs video cameras and a small set of
retro-reflective markers to create a low-cost, easy-to-use system that might
someday be practical for home use. The low-dimensional control signals from the
user's performance are supplemented by a database of pre-recorded human motion.
At run time, the system automatically learns a series of local models from a
set of motion capture examples that are a close match to the marker locations
captured by the cameras. These local models are then used to reconstruct the
motion of the user as a full-body animation.
We demonstrate the power and flexibility of this approach by having users
control six behaviors in real time without significant latency: walking,
running, hopping, jumping, boxing, and Kendo (Japanese sword art). The
reconstructed motion is based on a single large human motion database. Our
experiments indicate that this approach scales well with the size and
heterogeneity of the database and is robust to variations in kinematics between
users. The resulting animation also captures the individual style of the user's
motion through spatial-temporal interpolation of the data. Finally, we assess
the quality of the reconstructed motion by comparing against ground truth data
simultaneously captured with a full marker set in a commercial motion capture
system.
Publications
·
Performance Animation from
Low-dimensional Control Signals, ACM Transactions on Graphics
(SIGGRAPH 2005)
PDF (1.6M); final video (83Mb mov
clip with audio)
Videos, data and slides
- Walking
from 6 Markers and two cameras
- Running from 6 markers and two cameras
- Jumping
from 6 markers and two cameras
- Leaping
from 6 markers and two cameras
- Boxing from 6 markers and two cameras
- Kendo from 6 markers and two cameras
- Transitions between different motion types
- Animations from one camera
- Testing on different users
Project members
Related
projects
- Vision-based
Control of 3D Facial Animation,
ACM SIGGRAPH/Eurographics Symposium on Computer Animation (SCA
2003)
- Interactive
Control of Avatars Animated with Human Motion Data, ACM
Transactions on Graphics (SIGGRAPH 2002)
Acknowledgement
Supported in part by the NSF under Grant
IIS-0205224 and IIS-0326322.