Structuralizing Multimedia Data
IEEE MultiMedia
ETP '03 Proceedings of the 2003 ACM SIGMM workshop on Experiential telepresence
Personal chronicling tools for enhancing information archival and collaboration in enterprises
Proceedings of the the 1st ACM workshop on Continuous archival and retrieval of personal experiences
Event-based modeling and processing of digital media
Proceedings of the 1st international workshop on Computer vision meets databases
Event-based multimedia chronicling systems
CARPE '05 Proceedings of the 2nd ACM workshop on Continuous archival and retrieval of personal experiences
| Hi-index | 0.00 |
A key challenge to the successful application of the data modeling, storage and retrieval is the processing of relevant media to determine the events that are represented by them. Towards this goal, we are developing a heterogeneous media events processing system that combines human expertise with algorithmic processing capabilities. In doing so, we seek to leverage the fact that computers are good at fast low level processing and humans are good at high level analysis and understanding. Our approach to bridging the "signal-to-symbol" barrier is based on a 3-layered event-processing architecture composed of a lower data event layer, higher domain event layer, and an elemental event layer in the middle, which links data and domains with symbolic indices (see [3] for details). Two important challenges in this context are providing humans with an interactive interface to do a domain level event tagging, and providing computers with methods to construct computational causality models based on tagged events. Towards this, in our demonstration, we first show how our system detects events and we provide user-tagging environment on top of these detected events. Later, we will introduce our distributed media processing networks and specific domain based event search system, which will show a way to change existing limited PowerPoint like presentation methods.