[PYTHON] The Python programming language and all associated documentation is available via anonymous ftp from: ftp.cwi.nl. [DIVER] R. Gossweiler, C.

Size: px

Start display at page:

Download "[PYTHON] The Python programming language and all associated documentation is available via anonymous ftp from: ftp.cwi.nl. [DIVER] R. Gossweiler, C."

Patrick Daniels
5 years ago
Views:

1 [PYTHON] The Python programming language and all associated documentation is available via anonymous ftp from: ftp.cwi.nl. [DIVER] R. Gossweiler, C. Long, S. Koga, R. Pausch. DIVER: A Distributed Virtual Environment Research Platform. IEEE [TCR] R. Gossweler, R. Pausch. A System for Application-Independent Time Critical Rendering. Short paper in CHI 94 proceeedings. [HEAD] R. Pausch, M. Shackelford, D. Proffitt. A User Study Comparing Head-Mounted and Stationary Displays. IEEE

2 Infrastructure While building two-handed interaction in Alice is relatively straight forward, Alice presently has no event model; application programmers poll the devices directly and perform their own pick correlation. Application programmers should not have to attend to these low-level details: it distracts them from the more important issues of building new applications and demonstrating new interaction techniques. Building a full multi-modal event-based input model (two hands, voice input, 3D sound, etc.) remains an open area of research. Programmers must also not be distracted by building and rebuilding low-level mechanisms required to maintain immersive update rates. Keeping the frame rate sufficiently high is a concern that all 3D interactive applications will have, and because scenes can become arbitrarily complex, this concern cannot ever be fully resolved by simply employing more powerful hardware. When the scene complexity rises beyond the capabilities of the hardware, the system can degrade either the frame rate or the image quality. While it is easier for current systems simply to degrade the frame rate, perceptual psychologists contend that this is an inappropriate compromise, and that there are techniques for degrading the image quality which exploit the human perceptual system. Collaboration Over the past two years, our collaboration with perceptual psychologist Dennis Proffitt at the University of Virginia has been extremely valuable, producing important results in both Computer Science and Psychology. Thanks to the help of Dr. Proffitt, we have the first known confirmation of a task in which a headtracked point of view outperforms a point of view controlled by the user s hand [HEAD]. In other research, Dr. Proffit worked with us in an experiment that measures people s ability to estimate the incline of hills. In this work, we were able to show that people overestimate the incline of virtual hills seen in a head mounted display (HMD) by the same amount as they do with hills in the real world. At least in this limited domain, 3D interactive immersive graphics proved to be a powerful lab apparatus for the perceptual psychologists. Clearly, this collaboration benefits both fields. Psychologists benefit by gaining powerful new tools and a new domain in which to conduct and extend their research. Computer scientists benefit from the expertise and experience of psychologists, who have already faced in the real world many of the issues we are now facing in our synthetic worlds. This exchange of ideas has lead to very concrete benefits: we are persuing patent protection of several new display methodologies based on insights from Dennis Proffitt. Our perceptual psychologist colleagues have, in our minds, increased the urgency of addressing the socalled registration problem. This is the general problem of ensuring that graphics in the HMD align correctly with objects in the real world (incliuding the user). Not solving this problem, (or failing to uderstand the scope of this problem) may limit the future utiliy of HMD s for perceptual psychology research and will hamper the progress of research in the area of Augmented Reality, where synthetic graphics are superimposed over a view of the real world. Closing Rapid prototyping tools and collaboration with members of the perceptual psychology community may help us take advantage of what to date has remained something the insurmountable opportunity of 3D interaction research. References [ALICE] M. Conway, R. Pausch, R. Gossweiler, T. Burnette. Alice: A Rapid Prototyping System for Building Virtual Environments. Short paper in CHI 94 proceeedings.

3 UVa Position Paper for CHI 94 Workshop on 3D Interaction Introduction Matthew Conway, Rich Gossweiler, and Randy Pausch Department of Computer Science, University of Virginia, Charlottesville, VA {conway rich pausch Three-dimensional user interface research represents a broad new domain. Working in this field is time consuming and laborious, in part, because there are so many possibilities to explore. Progress remains slow, unfortunately, because the development tools are still primitive and because most systems are still hampered by the edit-test-compile cycle imposed by C and C++. We must have tools which allow rapid prototyping so that we can quickly build and experiment with new interaction ideas. As long as the tools impose long delays between design iterations, our output will remain small and good ideas will remain untried. Simply being able to iterate quickly over designs is not sufficient. Blindly trying one thing after another is not likely to yield better research. In a domain as large as this, instead of using our own intuition to guide our research, we need to embrace the expertise of researchers in other fields. Just as 2D GUI researchers reached out to graphics designers and anthropologists, we should reach now to architects and perceptual psychologists, who have important insights into the questions we are asking regarding the ways in which people percieve space and how they interact with the objects in space around them. These experts will help guide our ideas and keep us from exploring blind avenues of research, and along the way, we will uncover ideas and phenomena new to both our fields. In fact, this has already happened. Development Tools Our rapid prototyping system, Alice [ALICE], is a distributed system whose design emphasizes quick turnaround time for the programmer while not sacrificing the update rates that the user requires in order to remain immersed. Alice gains its flexibility via an object oriented, interpreted language called Python, which was developed by Guido Van Rossum at CWI [PYTHON]. In this system, the rendering engine and the simulation engine run asynchronously on different machines, so that the update of the user s view point and the feedback of the user s hands remain unaffected by heavy loads imposed by any computation that the application may be performing. Because Alice is object oriented, we find that we can encapsulate powerful abstractions in easy-to-use reusable modules and classes. For example: researchers from Brown were able to assemble a new twohanded ladder climbing interaction after their first two sessions with Alice, using a glove class developed previously for another application. We have included for reference a videotape detailing some of Alice s features that we included in our recent ARPA funding proposal.

4 [PYTHON] The Python programming language and all associated documentation is available via anonymous ftp from: ftp.cwi.nl. [DIVER] R. Gossweiler, C. Long, S. Koga, R. Pausch. DIVER: A Distributed Virtual Environment Research Platform. IEEE [TCR] R. Gossweler, R. Pausch. A System for Application-Independent Time Critical Rendering. Short paper in CHI 94 proceeedings. [HEAD] R. Pausch, M. Shackelford, D. Proffitt. A User Study Comparing Head-Mounted and Stationary Displays. IEEE

5 Infrastructure While building two-handed interaction in Alice is relatively straight forward, Alice presently has no event model; application programmers poll the devices directly and perform their own pick correlation. Application programmers should not have to attend to these low-level details: it distracts them from the more important issues of building new applications and demonstrating new interaction techniques. Building a full multi-modal event-based input model (two hands, voice input, 3D sound, etc.) remains an open area of research. Programmers must also not be distracted by building and rebuilding low-level mechanisms required to maintain immersive update rates. Keeping the frame rate sufficiently high is a concern that all 3D interactive applications will have, and because scenes can become arbitrarily complex, this concern cannot ever be fully resolved by simply employing more powerful hardware. When the scene complexity rises beyond the capabilities of the hardware, the system can degrade either the frame rate or the image quality. While it is easier for current systems simply to degrade the frame rate, perceptual psychologists contend that this is an inappropriate compromise, and that there are techniques for degrading the image quality which exploit the human perceptual system. Collaboration Over the past two years, our collaboration with perceptual psychologist Dennis Proffitt at the University of Virginia has been extremely valuable, producing important results in both Computer Science and Psychology. Thanks to the help of Dr. Proffitt, we have the first known confirmation of a task in which a headtracked point of view outperforms a point of view controlled by the user s hand [HEAD]. In other research, Dr. Proffit worked with us in an experiment that measures people s ability to estimate the incline of hills. In this work, we were able to show that people overestimate the incline of virtual hills seen in a head mounted display (HMD) by the same amount as they do with hills in the real world. At least in this limited domain, 3D interactive immersive graphics proved to be a powerful lab apparatus for the perceptual psychologists. Clearly, this collaboration benefits both fields. Psychologists benefit by gaining powerful new tools and a new domain in which to conduct and extend their research. Computer scientists benefit from the expertise and experience of psychologists, who have already faced in the real world many of the issues we are now facing in our synthetic worlds. This exchange of ideas has lead to very concrete benefits: we are persuing patent protection of several new display methodologies based on insights from Dennis Proffitt. Our perceptual psychologist colleagues have, in our minds, increased the urgency of addressing the socalled registration problem. This is the general problem of ensuring that graphics in the HMD align correctly with objects in the real world (incliuding the user). Not solving this problem, (or failing to uderstand the scope of this problem) may limit the future utiliy of HMD s for perceptual psychology research and will hamper the progress of research in the area of Augmented Reality, where synthetic graphics are superimposed over a view of the real world. Closing Rapid prototyping tools and collaboration with members of the perceptual psychology community may help us take advantage of what to date has remained something the insurmountable opportunity of 3D interaction research. References [ALICE] M. Conway, R. Pausch, R. Gossweiler, T. Burnette. Alice: A Rapid Prototyping System for Building Virtual Environments. Short paper in CHI 94 proceeedings.

6 UVa Position Paper for CHI 94 Workshop on 3D Interaction Introduction Matthew Conway, Rich Gossweiler, and Randy Pausch Department of Computer Science, University of Virginia, Charlottesville, VA {conway rich pausch Three-dimensional user interface research represents a broad new domain. Working in this field is time consuming and laborious, in part, because there are so many possibilities to explore. Progress remains slow, unfortunately, because the development tools are still primitive and because most systems are still hampered by the edit-test-compile cycle imposed by C and C++. We must have tools which allow rapid prototyping so that we can quickly build and experiment with new interaction ideas. As long as the tools impose long delays between design iterations, our output will remain small and good ideas will remain untried. Simply being able to iterate quickly over designs is not sufficient. Blindly trying one thing after another is not likely to yield better research. In a domain as large as this, instead of using our own intuition to guide our research, we need to embrace the expertise of researchers in other fields. Just as 2D GUI researchers reached out to graphics designers and anthropologists, we should reach now to architects and perceptual psychologists, who have important insights into the questions we are asking regarding the ways in which people percieve space and how they interact with the objects in space around them. These experts will help guide our ideas and keep us from exploring blind avenues of research, and along the way, we will uncover ideas and phenomena new to both our fields. In fact, this has already happened. Development Tools Our rapid prototyping system, Alice [ALICE], is a distributed system whose design emphasizes quick turnaround time for the programmer while not sacrificing the update rates that the user requires in order to remain immersed. Alice gains its flexibility via an object oriented, interpreted language called Python, which was developed by Guido Van Rossum at CWI [PYTHON]. In this system, the rendering engine and the simulation engine run asynchronously on different machines, so that the update of the user s view point and the feedback of the user s hands remain unaffected by heavy loads imposed by any computation that the application may be performing. Because Alice is object oriented, we find that we can encapsulate powerful abstractions in easy-to-use reusable modules and classes. For example: researchers from Brown were able to assemble a new twohanded ladder climbing interaction after their first two sessions with Alice, using a glove class developed previously for another application. We have included for reference a videotape detailing some of Alice s features that we included in our recent ARPA funding proposal.

preface Motivation Figure 1. Reality-virtuality continuum (Milgram & Kishino, 1994) Mixed.Reality Augmented. Virtuality Real...

preface Motivation Figure 1. Reality-virtuality continuum (Milgram & Kishino, 1994) Mixed.Reality Augmented. Virtuality Real... v preface Motivation Augmented reality (AR) research aims to develop technologies that allow the real-time fusion of computer-generated digital content with the real world. Unlike virtual reality (VR)