C2Nav: Collaborative Constrained Navigation Project

Project Description

Navigating through the large virtual worlds or complex data visualizations typical of virtual reality applications can be quite difficult. Frequently, there is a poor match between the goal of such a navigation activity, the control device, and the skills of the average user. With the recent rise in interest in collaborative virtual environments, we see the need for systematic attention to intelligent focussing of the attention of collaborative users toward objects and artifacts of mutual concern. The collaborative navigation task becomes even more challenging when users must balance the cognitive demands of personal exploration and audio communication while attempting to mimic collaborators' viewpoints or maintain visual contact with collaborators' avatars. We are exploring a range of techniques for facilitating cooperative navigation of these shared virtual environments. Our objective is to diminish the cognitive load on the participants, thereby freeing them to focus on understanding the content of the virtual world.

We define the dual goals of collaborative virtual environments as facilitating information and viewpoint sharing while simultaneously promoting individual exploration and personal insight. These goals are valid for situations where the participants are interacting in either an expert-student or a peer-peer relationship. Within this context, we explore a gamut of methods bounded by strict viewpoint synchronization on one end to complete viewpoint independence on the other end. Specific methods include slow-out/slow-in synchronization on demand, fixed position and orientation tethering, and elastic position and orientation tethering. The application of 2D navigation manifolds provides a number of additional benefits including "hard" or "soft" collision detection, pre-determined view or vista points, autonomous "guided-tour streams", and overall facilitation of navigation for the users.

This research is significant for collaborative environments as it will lead to a set of tools and a software library which will "hide the details" of effective navigation from the application developers in much the same way that CAVERNsoft hides networking details and the CAVE library hides VR system details.

Project Participants

Constrained Navigation for Desktop Viewing with 2D Controllers - Examples from IEEE Visualization '97 paper

The set of images below show three different 2D constraint grids and their effects on a given sequence of mouse input values (specified by the orange path).

* Planar Constraint Surface

* Fixed Camera Orientation

* Planar Constraint Surface

* Modulated Camera Orientation

* Complex Constraint Surface

* Modulated Camera Orientation

The grid designer may include virtual vista points to give the user overviews of the entire environment or close-ups of a specific area.

Virtual Overlook
Virtual Point of Interest

The techniques may be applied to a wide variety of VR and visualization applications. Below is an example of a constraint field being used for molecular structure visualization.

* Camera constrained to surface
of torus

* Modulated camera orientation

Click on each image for the
associated MPEG (~2 Mb each)

Examples from early, non-collaborative CAVE prototypes

A terrain model with a family of related 2D constraint surfaces; these implicitly form a 3D manifold that can be accessed by a 3D motion of the controller. Also note that, in principle, this could be an exploded schematic of set of 2D surfaces that coincided at many points in 3D space instead of being separated as they are here for clarity.
A clock tower and a cylindrical guide manifold de-signed to inspect it from all angles using a simple 2D plane of navigation parameters. The PVE (Physical Viewing Environment) is shown as a white box, with the user's head symbolized as a red globe.
Exploring a large molecule using the constrained navigation paradigm. Shown is an overview of the guide manifold, sample camera models, and a representative position of the PVE shown as a white box.
The molecule scene as viewed from inside the immersive environment.