The goal of this project is to foster new research collaborations across different disciplines by developing a novel massive interaction game that represents a unique genre wherein teams of users strategically direct armies of virtual 3D "terra cotta" soldiers by employing physical drum beats as control signals to determine the warriors' behavior. It is intended that the game will engage and enthrall both user teams and a live audience with the special demands of cooperative and competitive game play in a compelling and creative scenario. Successful completion of the research will require intensive collaboration across multiple fields of computational science, art, and music. The culmination of the project will be a social event centrally located on the Virginia Tech campus, in which the finished game is publically introduced to the community at large.

To achieve these goals, the PI team will engage interactive game design approaches to draft and enrich the overall game experience. They will adopt commoditized Graphics Processing Units (GPUs) to meet the computational challenges of real-time simulation, animation, and rendering of massive number of 3D characters. Novel numerical methods will be developed to enhance the efficiency of global path finding for massive crowd simulation. The individual combat behavior of the virtual soldiers will be automatically regulated by an artificially intelligent component, which is supported by GPU accelerated agent-based simulation. In order to provide visualization capacity for a massive crowd in an enormous environment, the investigators will develop technologies to enable level-of-detail rendering and animation. For massive interaction support, signal processing and computer vision approaches will be used to recognize the drumbeats and to interpret the audience participation behavior (which will be tracked with the aid of microphones embedded among the members of audience and cameras aimed at them). Synthetic music generation approaches will be utilized to produce real-time accompaniment to the drums.

Broader Impacts: The outcome outlined above will be the result of broad and intensive collaboration between engineering and the arts. Computer engineers will design and develop the overall game engine architecture, graphics rendering technologies, numerical solutions for character path planning, and approaches to recognize drum beats. Graphics artists will design 3D geometry models and animations for "terra cotta" soldiers, and will work closely with engineers to build a smooth "pipeline" to import digital content into the game software system. Researchers in modern music will produce accompaniment to the cadence of the drums. The team-based approach adopted in this project will expose students to cross-disciplinary research, so that they come away with an understanding and appreciation of the need to function cooperatively to achieve rewarding and exciting results. Teaming of undergraduate students with graduate students will, it is hoped, lead some of the undergraduates to choose research related career paths while also encouraging the graduate students to learn how to better mentor others. The PI team hopes that the mass media event at the culmination of the project will imbue the broader public (and especially K-12 students) with the excitement of computing and the arts.

Agency
National Science Foundation (NSF)
Institute
Division of Information and Intelligent Systems (IIS)
Type
Standard Grant (Standard)
Application #
0940723
Program Officer
Ephraim P. Glinert
Project Start
Project End
Budget Start
2009-07-01
Budget End
2012-09-30
Support Year
Fiscal Year
2009
Total Cost
$149,648
Indirect Cost
City
Blacksburg
State
VA
Country
United States
Zip Code
24061