|
Click on a theme or a project in the table below for more information.
Project leader:
Ir. Frits Post (TUD)
Consortium:
UU, TUD, CWI
Industrial partners (non-exhaustive):
KINEO CAM, CycloMedia, PS-Tech
Total FTE: 4.30 (heads: faculty: 8, PhD: 3)
Key BRICKS publications:
| • |
R. Geraerts, M.H. Overmars: "A General Framework for Real-Time High-Quality Path Planning" In: Computer Animation and Virtual Worlds, 2007, to appear
|
| • |
D. Nieuwenhuisen, A.F. van der Stappen, M.H. Overmars:"An effective framework for path planning amidst movable obstacles" In: Proceedings Workshop on Algorithmic Foundations of Robotics, 2006)
|
| • |
G. de Haan, M. Koutek, and F. Post:"IntenSelect: Using dynamic object rating for assisting 3D object selection" In: Proceedings of the 9th IPT and 11th Eurographics VE Workshop (EGVE) '05, pp. 201-209, 2005
|
| • |
G. de Haan, M. Koutek, and F. Post: "Flexible abstraction layers for VR application development" In: Proceedings of IEEE Virtual Reality 2007, pp. 239-242
|
| • |
J.D. Mulder: "Occlusion in mirror-based co-located augmented reality systems" In: Presence: Teleoperators and Virtual Environments 15, 2006, pp. 93-107
|
|
Project MSV2: Interactive Virtual Environments
The overall goal of this project is to create interaction
techniques for virtual environments that are natural and
effective. When a user interacts with a virtual environment
one can distinguish three levels of tasks. On the highest
level the user tries to achieve certain (strategic) goals.
For example, in a safety training simulation the user may try
to rescue a victim. On the intermediate level the user breaks
these tasks up into certain actions, like walking to the
victim or picking up some object. On the lowest level these
actions are performed by the user with motions of the
joystick, mouse, or other devices that directly steer the
camera, avatar or some object in the environment. Most of
today's systems require the user to perform these low-level
actions directly. For example, users often need to navigate
through the environment. In most of the present systems this
requires direct control by the user (through arrow keys,
mouse and/or a joystick) and especially for inexperienced
users this is difficult and requires a lot of attention,
distracting them from the higher-level tasks they wish to
accomplish. Effective virtual environments must enable the
user to concentrate on the higher-level tasks at hand. We
need a more intelligent interaction system that facilitates
the higher-level tasks by providing better (more intuitive)
low-level interaction tools and, preferably, by automating
most of the intermediate-level actions.
Industrial cooperation
- Together with the company CycloMedia we are working on interfaces with a virtual world that is constructed from 360 degree panoramic images that CycloMedia takes every 10 meters on the Dutch road network. The interface will use navigation techniques developed in MSV2. This project, called CycloCity, is also partially funded by the BSIK project "Ruimte voor GeoInformatie".
- We also cooperate with TNO Science and Industry to improve operator support for large scale camera surveillance control rooms. Here we investigate the use of 3D information displays and 3D navigation techniques.
- Further, we established mutual working relations with the company Personal Space Technologies that commercializes the personal space station, developed at CWI. This has resulted in a SENTER funded project.
International cooperation
The project partners have worked together with many groups
around the world, e.g. with groups in LAAS and TelAviv on
path planning in games and with Berkeley on manipulation.
Highlights 2004-2006
The MSV2 project has led to an increased interest in the
Netherlands for this area. The partners were instrumental in
the new NWO program VIEW and Utrecht obtained a FES-grant on
Gaming and Simulation that will bring together many
researchers and companies.
Research highlights
- We have developed a new approach to navigation and path planning based on the notion of corridors. This approach combines high-level planning with low-level flexibility leading to a generic, robust and very efficient technique that can easily be used in interactive environments, like games.
- In May 2007 Dennis Nieuwenhuisen has defended his PhD thesis on Path planning in changeable environments.
- We also developed a novel interaction technique which assists users in the hard task of handling small or moving objects in Virtual Environments, such as interface components. A probability scoring metric determines, based on users dynamic hand movements, which object a user intends to select. We incorporate this in a new software architecture for the rapid prototyping of domain specific VR applications.
- We have contributed to the development of a prototype for a new projector-based desktop VR workstation P-DRIVE.
- Also, we developed new techniques for providing consistent viewing and interaction to multiple cooperating users in a virtual environment.
Economic & societal impact
As indicated above, many companies are interested in the
project results and it is expected that these will be
incorporated in some products during the rest of the project.
In 2006, the CWI spin-off company PS-Tech (www.ps-tech.nl) was created
based on results partially developed in the MSV2 project.
Future work 2007-2009
- We will continue our work on navigation and plan to develop our corridor method into a generic, easy-to-use library. Industry is very interested in that.
- Also we will use our new motion-capture system to better understand the notion of natural navigation and manipulation.
- We will also continue our work on a prototyping framework for 3D interaction techniques, which will allow interaction designers to quickly design, develop, and evaluate many variants of domain specific, collaborative interaction scenarios for Virtual Environments.
- We will also contribute to a project on local multi-user interaction with virtual environments, and we will explore the use of 3D virtualenvironments for viewing and navigation through multiple surveillance video streams for improving situational awareness.
- Finally, we will research methods for quantative analysis of interactive tasks in order to provide more robust guidelines for designers of spatial interactive systems.
MSV2 Researchers funded by BRICKS
- Ir. F.H. Post (TUD)
- Ir. G. de Haan (TUD)
- Dr.ir. M. Koutek (TUD)
- Prof.dr. M.H. Overmars (UU)
- Dr. A.F. van der Stappen (UU)
- Dr. R. van Oostrum (UU)
- Dr. R. Geraerts (UU)
- Prof.dr.ir. R. van Liere (CWI)
- (vacancy postdoc) (CWI)
For more information, please refer to the publications and posters of this project.
|
