1. Executive Summary

The five sites that make up the Graphics and Visualization Center embarked six years ago on a concentrated, long-term, interdisciplinary research effort in computer graphics and scientific visualization. The sites joined in a Center because of their complementary expertise and their desire to collectively bring a new and broader perspective to research in graphics and visualization, one that could not be achieved by a single site. Together, we work to realize the extraordinary potential of computer graphics: flexible, robust, scientifically grounded methods of visual communication, analysis, and design.

Computation is an essential part of scientific inquiry, both for making scientific predictions and for experimentation. We work to make simulations that don't just ``look good'' but ``are good.'' We have moved from making attractive pictures to producing intuitively controlled simulations modeled with physically based constraints that can produce visual images and physical prototypes with known accuracy. The implications of these advances will reach far beyond computer graphics, for example, to enhance medical diagnosis and treatment, shorten design cycles, and predict the behavior of complex devices in as-yet unbuilt environments.

The following long-term research goals further illustrate our vision:

• Goal: Geographically dispersed multidisciplinary design and manufacturing teams should be supported by powerful, easy-to-use, intuitive tools for collaboratively sketching, visualizing, modifying, and prototyping mechanical designs. These tools should be integrated into a system that gives participants a visceral sense of working together in a shared space.
  Achieved: A testbed telecollaboration system for solving real-world manufacturing problems by creating experimental prototypes needed for Center research. The system includes: a shared virtual-reality application (VRAPP) with view-dependent avatars, integrated multi-user support for the Alpha_1 parametric 3D geometric modeling system, and interaction techniques that address problems unique to collaborative environments for 3D design.
• Goal: Models should be capable of representing increasingly complex shapes and manufacturing processes
  Achieved: New, more powerful representations of 3D geometry, such as manifolds, second-generation wavelets, and torn B-spline surfaces.
• Goal: Fully natural, task-appropriate 3D interaction should be the norm, not the exception.
  Achieved: A new gestural approach to 3D geometric modeling, which includes two-handed natural gestural drawing idioms linked for direct input to parametric 3D geometric modeling systems.
• Goal: Numerous application areas should be supported by the accurate computation and display of visual data and the integration of this data with other media. For example, architects and other designers should be able to walk through photorealistically rendered models of complex and dynamic synthetic environments that are merged with image streams of real people and objects. In addition, users should be able to make changes to these models interactively.
  Achieved: More accurate physical and perceptual lighting techniques, including a new class of primitive functions with nonlinear parameters for representing reflectance functions.
• Goal: Wide-area networked augmented-reality applications should be feasible. For example, physicians should be able to explore accurate visualizations of patient data (such as MRI scans) with augmented-reality displays in real-time and should be able to share such information with consulting physicians in distant facilities.
  Achieved: A new compact video-based see-through head-mounted display (HMD) that gives the user unprecedented visual integration of images on the display surfaces with objects in the real world. The HMD is being used in test scenarios in which ultrasound data is merged with video to give the physician an accurate view inside the patient during biopsy procedures. This new HMD was one of the 11 specific projects highlighted in NSF's 1998 budget request to Congress.
• Goal: Biologically based models should go beyond geometric specification to incorporate behaviors such as growth, movement, and interaction with their environments.
  Achieved: A new type of computer graphics modeling based on multicellular development. Using structured modeling techniques, our developmental models combine elements of the chemical, cell lineage, and mechanical models of morphogenesis.

The Center's ability to mount coordinated efforts that draw on many subdisciplines within and beyond computer science makes it possible to tackle ambitious research projects beyond the capacities of single-site labs, particularly the foundational projects needed for advances in graphics and visualization.

The Center has also made possible a sustained commitment to education and outreach with the dual goals of training the next generation of researchers and attracting new and diverse participants. Center faculty are deeply involved in teaching both undergraduate and graduate students in our labs, and the Center's education and outreach programs affect our local communities as well as the computer graphics field as a whole. We are now concentrating on leveraging our outreach by providing programs for teachers and by creating Web-based materials to reach large numbers of teachers and students.

Our linkages to industry remain strong, not in the least through a steady flow of graduates into both established companies and startups. Companies face a shortage of qualified graphics specialists and recruit proactively on our campuses. Despite major downsizing of research in the workstation industry, we have fostered relationships with industrial sponsors that include direct funding, technical exchanges and collaborations, personnel exchanges, and donations of lab equipment. Center alumni frequently maintain ties that add value to our students' educational experience and ensure a continuous dialogue between industrial and academic interests.

The Future

The commercial success of computer graphics as a new medium for creation, communication and analysis comes directly out of fundamental research done in the last decade. It is imperative, however, that this success not distract attention from the need to continue the underlying research. By continuing the high level of research for which the Center is known, we remain in a unique position to fulfill our goal of working together to strengthen the foundations of computer graphics to help the field realize its potential.

The Center's already strong impact on industry will facilitate our plans for additional industry linkages. Our high profile in education will be supported by continued integration of research with education, and the creation of interactive curricular resources will impact thousands of teachers and students. Also, our expanding commitment to programs for women and minorities will continue at all levels.

Conclusion

The success of computer graphics to date has proven its potential but in many ways it threatens its own progress. Mounting commercial and academic pressures force companies to concentrate on ever shorter product cycles and incremental improvements. In contrast, our Center is able to emphasize long-term, fundamental research that spans the discipline, has strong interdisciplinary components, and benefits from scientific and educational partnerships with many other institutions. A multi-site Center committed to the long-term goals of advancing the foundations of computer graphics is now more important than ever.