Tutorial-1: 3D Virtual Worlds: Challenges, Opportunities and Current Research

Duration: 2 Hours

Rich, large-scale, user-generated virtual worlds have been a part of science fiction for decades. Systems such as Second Life, Eve Online and World of Warcraft have been able to bring this vision to reality to some extent. However, such worlds still present significant technical challenges due to the limitations of computational and network resources. A virtual world can be larger than what a single machine will store, so it is distributed over a set of cloud servers and has to be loaded over the Internet. Furthermore, since the world may contain millions of 3D objects, it is too expensive for a graphics card to render them at interactive frame rates. Finally, due to their dynamic, user-generated nature, we cannot simply rely on hand-crafted or pre-computed levels of detail. Combined, these factors lead to several research problems, including how to dynamically partition a world across multiple servers, and how to automatically compact, aggregate and simplify 3D content.

Initially, my talk will focus on an introduction and overview of existing virtual world systems. It will then describe some of the challenges and limitations that still exist in today’s virtual worlds. Finally, it will focus on the design and implementation of a research-based virtual world system that overcomes many of the limitations of existing virtual worlds. This includes a simplification pipeline for 3D models in virtual worlds. The pipeline addresses the problems described above using three novel algorithms, one to group together visually similar objects, one for deduplicating similar objects within these groups, and one for simplifying the resulting aggregation of objects. Together, these algorithms leverage the spatial coherence and instancing typical to virtual worlds to enable large, immersive environments consisting of hundreds of thousands of objects to quickly load from the cloud and render at interactive frame rates.
 

Key objectives

- To get an introduction and overview of 3D virtual worlds.
- To explore the challenges in the field of virtual world systems.
- To get an introduction to state-of-the-art research in virtual worlds.

Tutorial Outline

- Introduction to 3D Virtual Worlds
- Challenges and limitations in existing virtual worlds
- Design and Implementation of Sirikata, a new research-based virtual world
- Design of a new simplification pipeline for virtual world content
- Opportunities for further research in this area
 

Targeted Audience

Our target ordinance includes academics and companies, postgrad students, and other researchers and engineers. The tutorial will be a great venture especially for students at both the graduate and under-graduate levels to pursue research/development in this area.

 

Tutorial Fees

For Students:
IEEE Student Members: Rs. 500
Non-IEEE Student Members: Rs. 800
For Professionals:
IEEE Members: Rs. 3000
Non-IEEE Members: Rs. 4000

Instructor Biography

Tahir Azim is currently an assistant professor at the National University of Sciences and Technology (NUST) in Pakistan. His research interests include distributed systems, networks, 3D virtual worlds, and parallel computing. He received a BE from NUST, and an MS and PhD from Stanford University in 2014 under the guidance of Philip Levis. In the past, he has also been actively involved in distributed systems research projects at the California Institute of Technology (Caltech) and the European Organization for Nuclear Research (CERN).