|
|
|
|
1 Non Photorealistic Rendering
Figure 1: Professor Goochs research combines computer graphics techniques for creating artistic imagery with the evaluation methods of perceptual psychology to optimize visual communication using images, animations, and interactive applications. Images one through three show examples of automatically generated technical illustrations. Images four though six show automatically generated cartoon images. In the computer graphics community, rendering is the process by which data is converted into images. Photorealistic rendering denotes images based on physical simulations. The goal of photorealistic rendering is to create images indistinguishable from photographs of real world scenes. In contrast, non-photorealistic rendering (NPR) is concerned with images that are guided by artistic processes. The reasons for using NPR range from aesthetic and stylistic considerations to communicative advantages such as abstraction and symbolism. In his extensive body of non-photorealistic rendering work, Professor Gooch has explored a variety of computer based methods for producing illustrations including: optimization for scene layout [6], using silhouette and boundary edges to enhance form [7, 8, 16], as well as stippling, paint-strokes, and textures to indicate material properties [9, 32]. Dr. Gooch has also done work in creating digital oil paintings [5, 15]. In computer graphics the way that light interacts with objects is abstracted mathematically into a shading model. There are four local shading models in computer graphics, each named for the investigator who discovered it; Phong, Gouraud, Blinn and Gooch [1]. The Gooch shading model, discovered by Dr. Gooch, is widely used in academia for scientific visualization [33], as well and in movies and video games. Professor Gooch has made contributions in color based image processing to enhance the appearance of digital images. He done work in color transfer between images [24] and in converting color images to grayscale in way that maintains color differences [4]. Dr. Gooch is the author of two books on NPR: Non-Photorealistic Rendering, Bruce Gooch and Amy Gooch published by A.K. Peters 2001; and Illustrative Visualization, Bruce Gooch, Amy Gooch and Mario Costa Sosa to be published by A.K. Peters in 2009. 2 Perception and Graphics
Figure 2: Images one through three show a photograph, illustration, and caricature. Behavioral studies, carried out by Dr. Gooch, demonstrate that learning tasks are faster with the illustrations or caricatures. His fMRI experiments demonstrate that more and different areas of the brain are active when decoding the the automatically generated facial illustrations than with photographs. Dr. Gooch introduced Semanticons, images four through six, to enhance the representation of files in a GUI by automatically generating icons that better reveal the meanings of a desktop files contents. Psycho-physical studies comparing Semanticons and standard desktop icons in visual search and memory tasks show Semanticons yield significant improvements in human performance. Professor Gooch seeks to understand the effectiveness of computer-generated imagery using perceptual metrics and evaluation techniques. Empirical studies repeatedly find that design evaluations based on simple introspection and preference are poor predictors of user performance. Instead, Dr. Gooch looks to the methodology of perceptual psychology to carefully design experiments that use task based performance to evaluate computer generated imagery. Professor Gooch is a pioneer in the perceptual evaluation of non-photorealistic imagery. He was the founding online editor of the ACM Transactions on Applied Perception. His research discoveries include finding increased learning speed using computer generated facial illustrations [7, 35], demonstrating enhanced depth perception in images via artistic matting [13], enhanced interfaces for detecting network attacks and misconfigurations [25], decision making in social contexts [14], detecting false captioning in images [13] and showing increased performance in memory and search tasks using modified GUI icons [27]. Dr. Gooch has also made contributions to computer graphics using computational models of the human visual system to solve known or novel problems. He has created methods for estimating light positions from digital photographs without the use of calibrated cameras [35]. Dr. GoochÕs introduced the use of computational models of the human visual system in creating three dimensional models from photographs or drawings [10, 26]. Dr. Gooch introduced techniques for automatically retargeting images and animations, that is, for adapting them for display at different sizes and aspect ratios while preserving the recognizability of important image features. The motivation for this work is the need for tools that allow us to author imagery once, and then automatically retarget that imagery for a variety of different display devices as needed. The problem is to retarget an image to a new size and/or aspect ration in a manner that preserves the recognizability. To do this, the features of the image must be identified, their importance determined, and then the image needs to be re-arranged such that the important features are well-represented. Professor Gooch introduced the retargeting problem to the computer graphics community in a 2004 ACM Siggraph presentation and has published work in image retargeting [30], video retargeting [29] and flash animation retargeting [31]. 3 Learning and Video Games
Figure 3: Dr. Gooch and his team have proposed and evaluated a methodologies and scaffolding for second language learning in massive multi-player online role-playing games (MMORPG). Right: Conversational Dialogue with Game Avatar. Left: Role-playing character chatting with non-role playing character in English. MMORPGs serve as the catalyst for fostering students language proficiency as students interact in a foreign language while playing the game. Learning is an active, personal experience that allows the student to reflect on what they know (e.g. beliefs, ideas, misconceptions, etc.) and how this knowledge shapes their understanding of the world and sense of self. Game-based learning refers to embedded instructional content in video games. Game-informed practices give students an opportunity to learn concepts in a situated manner. For example, students who play video sports games (i.e. football) learn about the rules of the game and the social practices (e.g. huddle to discuss strategy) associated with the game. Using a subset of game informed practices, Dr. Gooch has been studying the effect of the Massive Multiplayer Online Role Playing Game (MMORPG) Ever Quest 2 for second language acquisition [17, 19, 21, 22]. Dr. Gooch has also made contributions in computer science education using video games for affective learning in first year computer science courses [18, 20, 23]. Professor Gooch is recognized as a leader in the study of video games and learning by his peers in academia. He was the general chair of the ACM conference Interactive Three Dimensional Graphics and Games (I3D) in 2007 and Papers Co-Chair in 2008. Of the over 1200 journals and conferences in computer science I3D has been ranked by Cite-Seer as number 25 based on the impact of the conference papers. Dr. Gooch has also been on the program committee for the Foundations of Digital Games conference for 3 years and is currently serving as the publicity chair. Dr. Gooch has also been recognized by industry as an innovator. He has received $160,000 in two research awards for the study of video games and learning from Microsoft Research. Last year he was invited to demonstrate his game technology at Microsoft Research TechFest. Microsoft Research TechFest is currently know as the top computer technology demonstration in the world. Dr. Gooch is one of only three academics invited to demonstrate at TechFest 2008, the other participants were Brown and Columbia Universities. 4 Computational Aesthetics
Figure 4: Inspired by Jackson Pollocks painting processes, Dr. Gooch lead a team that developed an interactive 3D painting system for creating digital abstract paintings. This figure shows a comparison between a painting by Jackson Pollock Number 13A, 1948.(Left) and one created by a novice user in 30 minutes with the interactive 3D fluid jet painting system (Right). Pollock created his abstract paintings by throwing thinned household paint to create guided, semi-random patterns on his canvas. Since current 3D fluid jet simulation techniques are too computationally expensive to be used in realtime applications, Dr. Gooch developed a stable model with reduced dimensions. A second algorithm generates the two dimensional splatter patterns of high-speed droplet impact. Our ability to convey complex meaning and emotion by the clever arrangement of symbols and signs is one of the most celebrated aspects of our human heritage. If Aesthetics is defined as the interaction between symbol and observer within a social context, then Computational Aesthetics is the empirically based field that examines applications of Aesthetics. Computational Aesthetics investigates the creation of tools that can enhance the expressive power of the fine and applied arts and furthers our understanding of aesthetic evaluation, perception, and meaning. Dr. GoochÕs vision is of tools that both engage and inform the artistic process. Recent mathematical analysis indicates that the fluid jet patterns of PollockÕs painting may be related to their fractal structure[12]. PollockÕs paintings contain self-similar patterns which contribute to the aesthetic quality of the work. Dr. GoochÕs Fluid jet painting system includes an evaluation tool which calculates the fractal dimensions of a userÕs painting [11]. Unlike real-world paintings, the digital system makes users aware of fractal properties interactively, and they can compare the fractal characteristics their work to that of PollockÕs easily. This system was demonstrated to the over 30,000 attendees at the Siggraph conference in 2006. It has been on exhibition at the Singapore Museum of Science. Professor Gooch has also made contributions to the area of Computational Aesthetics research with his work on the artistic composition of models in a computer graphics scene [66], and on using artistic matting to increase perceived depth in digital images [2, 3]. His most interesting contribution may be his gaze dependent facial expressions system [28]. Based on artistic methods used for manipulating perception, Dr. Gooch presented a technique that creates facial images with conflicting emotional states at different spatial frequencies. Professor Gooch is a founder of Computational Aesthetics research. He was an organizer of the first ever Eurographics sponsored conference in the area (Girona Spain 2005), a Dagstuhl seminar (Germany 2006), the second conference at the Banff Centre (Canada 2007), and the third Eurographics sponsored conference (Lisbon Portugal 2008). Dr. Gooch also served as a guest editor of the IEEE journal ÓComputer Graphics and ApplicationsÓ special issue on Computational Aesthetics in 2007. The forth Eurographics sponsored conference on Computational Aesthetics will be held in Victoria Canada in 2009. ReferencesDigital versions of all of these papers are available at Dr. Goochs Publications Web Page.
|
| ||||||||
| ||||||||
| ||||||||
| ||||||||
| ||||||||
|
