We propose a new approach that utilizes semantic information to register 2D monocular video frames to the world using 3D georeferenced data, for augmented reality driving applications. The geo-registration process uses our predicted vehicle pose to generate a rendered depth map for each frame, allowing 3D graphics to be convincingly blended with the real world view. We also estimate absolute depth values for dynamic objects, up to 120 meters, based on the rendered depth map and update the rendered depth map to reflect scene changes over time. This process also creates opportunistic global heading measurements, which are fused with other sensors, to improve estimates of the 6 degrees-of- freedom global pose of the vehicle over state-of-the-art outdoor augmented reality systems [5]-, [19]. We evaluate the navigation accuracy and depth map quality of our system on a driving vehicle within various large-scale environments for producing realistic augmentations.

When directly manipulating 3D objects in an immersive environment we cannot normally achieve the accuracy and control that we have in the real world. This reduced accuracy stems from hand instability. We present PRISM, which dynamically adjusts the C/D ratio between the hand and the controlled object to provide increased control when moving slowly and direct, unconstrained interaction when moving rapidly. We describe PRISM object translation and rotation and present user studies demonstrating their effectiveness. In addition, we describe a PRISM-enhanced version of ray casting which is shown to increase the speed and accuracy of object selection.

We present a novel text input interface for immersive virtual environments called CTD, or "Connect the Dots". The CTD interface is a small virtual panel containing a grid of dots the user can connect to form alphanumeric characters using a hand held stylus. Coupled with a physical paddle or desk for force feedback, the CTD provides an intuitive text input interface similar to using a pen and paper.

A significant benefit of an immersive virtual environment is that it provides users the ability to interact with objects in a very natural, direct way; often realized by using a tracked, hand-held wand or stylus to grab and position objects. In the absence of force feedback or props, it is difficult and frustrating for users to move their arms, hands, or fingers to precise positions in 3D space, and more difficult to hold them at a constant position, or to move them in a uniform direction over time. The imprecision of user interaction in virtual environments is a fundamental problem that limits the complexity of the environment the user can interact with directly. We present PRISM (precise and rapid interaction through scaled manipulation), a novel interaction technique which acts on the user's behavior in the environment to determine whether they have precise or imprecise goals in mind. When precision is desired, PRISM dynamically adjusts the control/ display ratio which determines the relationship between physical hand movements and the motion of the controlled virtual object, making it less sensitive to the user's hand movement. In contrast to techniques like Go-Go, which scale up hand movement to allow long distance manipulation; PRISM scales the hand movement down to increase precision. We present the results of a user study which shows that PRISM significantly out-performs the more traditional direct manipulation approach.

Shared simple virtual environment (SSVE) is an object-oriented software framework designed for object-focused small group collaboration in highly dynamic virtual environments. Compared to most other collaborative virtual environment systems, SSVE allows a team of six to eight people to collaborate closely on a highly interactive model construction task. SSVE is a "multi-multi" framework - multi-user, multi-threaded, multi-platform, and multi-modal. Numerous interaction, awareness and collaboration mechanisms are provided by the framework and by applications that use the framework, including "what you see is what I see," deictic references, and exclusive object access. Several concurrency control mechanisms are also supported through the framework, including those suitable for closely-coupled synchronous collaboration in highly dynamic virtual worlds. In addition, support is provided for emergent group collaboration modes, and a novel user interface widget, the collaboration tree, is introduced to facilitate subgroup formation and integration of task results. In short, SSVE is a framework for the development of applications which explore group collaboration mechanisms in highly-interactive shared virtual environments. A small pilot study that used the framework to validate the utility and stability of the framework, as well as to investigate larger issues of group collaboration, is also presented.

Modern interaction systems are usually event-driven. New input devices often require new event types, and handling input from the user becomes increasingly more complex. Frequently, the WIMP (Windows, Icons, Menus, Pointer) paradigm widely used today is not suitable for interactive applications, such a virtual reality applications, that use more than the standard mouse and keyboard input devices.In this paper, we present the design and implementation of the Dynamic Event Model for Interactive System (DEMIS). DEMIS is a middleware between the operating system and the application that supports various input device events while using generic event recognition to detect composite events.

No abstract available.

The promise of immersive virtual environment (VE) applications has been that they can make interaction with computer information and processes easier by providing a medium that more closely matches the user's real environment. Meeting this promise, however; has proven to be difficult, as the interface between the user and the computer is not complete, the metaphors for interacting with information are not always obvious, and the tools for including current interaction techniques are not sufficient to the task. We present SVIFT, the Simple Virtual Interactor Framework and Toolkit, which supports efforts to meet the interaction needs of immersive VE applications. SVIFT allows for the design and implementation of various interaction techniques that can be easily incorporated into many VE applications and combined with other interaction techniques to produce more complex interactions. We also discuss the differences between desktop and immersive environment interaction, and the implications of those differences on the design of an interactor framework.

We present a careful evaluation of the sensory characteristics of the CyberGlove model CG1801 whole-hand input device. In particular, we conducted an experimental study that investigated the level of sensitivity of the sensors, their performance in recognizing angles, and factors that affected accuracy of recognition of flexion measurements. Among our results, we show that hand size differences among the subjects of the study did not have a statistical effect on the accuracy of the device. We also analyzed the effect of different software calibration approaches on accuracy of the sensors.