Semantic-Aware Content Delivery Framework for 3D Tele-Immersion
Supervisor(s) and Committee member(s): Klara Nahrstedt (advisor), Roy Campbell (opponent), Indranil Gupta (opponent), Cheng-Hsin Hsu (opponent)
3D Tele-immersion (3DTI) technology allows full-body, multimodal interaction among geographically dispersed users, which opens a variety of possibilities in cyber collaborative applications such as art performance, exergaming, and physical rehabilitation. However, with its great potential, the resource and quality demands of 3DTI rise inevitably, especially when some advanced applications target resource-limited computing environments with stringent scalability demands. Under these circumstances, the tradeoffs between 1) resource requirements, 2) content complexity, and 3) user satisfaction in delivery of 3DTI services are magnified.
In this dissertation, we argue that these tradeoffs of 3DTI systems are actually avoidable when the underlying delivery framework of 3DTI takes the semantic information into consideration. We introduce the concept of semantic information into 3DTI, which encompasses information about the three factors: environment, activity, and user role in 3DTI applications. With semantic information, 3DTI systems are able to 1) identify the characteristics of its computing environment to allocate computing power and bandwidth to delivery of prioritized contents, 2) pinpoint and discard the dispensable content in activity capturing according to properties of target application, and 3) differentiate contents by their contributions on fulfilling the objectives and expectation of user’s role in the application so that the adaptation module can allocate resource budget accordingly. With these capabilities we can change the tradeoffs into synergy between resource requirements, content complexity, and user satisfaction.
We implement semantics-aware 3DTI systems to verify the performance gain on the three phases in 3DTI systems’ delivery chain: capturing phase, dissemination phase, and receiving phase. By introducing semantics information to distinct 3DTI systems, the efficiency improvements brought by our semantics-aware content delivery framework are validated under different application requirements, different scalability bottlenecks, and different user and application models.
To sum up, in this dissertation we aim to change the tradeoff between requirements, complexity, and satisfaction in 3DTI services by exploiting the semantic information about the computing environment, the activity, and the user role upon the underlying delivery systems of 3DTI. The devised mechanisms will enhance the efficiency of 3DTI systems targeting on serving different purposes and 3DTI applications with different computation and scalability requirements.
The Multimedia Operating Systems and Networking (MONET) Research Group, led by Professor Klara Nahrstedt in the Department of Computer Science at the University of Illinois at Urbana-Champaign, is engaged in research in various areas of distributed multimedia systems.