Video Analysis and Content Extraction (VACE) is a research program now under Disruptive Technology Office for an aggressive research program which is now in its third phase.
The Government, through the VACE Program, is seeking proposals for innovative, creative, high-risk research to achieve significant advancements in technologies and methods for ingesting, indexing, managing, accessing and understanding a large video corpus from multiple heterogeneous video data sources. It is the intent of this research to continue to advance the state-of-the-art in technologies and methods for advanced, automated and automatic video content extraction, intelligent content services, and underlying enabling technologies.
This is definitely a great area of interest to many segments of society, not just the Government. In the first two phases, this effort has resulted in more research in video analysis as practiced in regular computer vision. I consider that the efforts in computer vision for â€˜ingesting, indexing, managing, accessing and understanding a large video corpus from multiple heterogeneous video data sourcesâ€™ should take a very different perspective than the routine research in analyzing video to detect objects and video events.
A very good thing in this call for proposal is the heavy emphasis on â€˜eventsâ€™ in all the phases of research and system implementation. In fact if really adopted, event-oriented perspective may result in really â€˜disruptiveâ€™ progress in this area. But this has to be â€˜event-orientsâ€™ perspective not just video-event-orientation. Ultimately video is just a means to the end of finding events in real world where the data about the real world is predominantly coming from video. Unfortunately, academic research takes reductionism to the extreme and makes the problem trivial in many cases. And that is not the problem of academics community, but of our society that sets the reward system that is based on counting publications not on solving problems. So, much research in video considers video analysis as the goal rather than as one of the information sources, obviously the richest information source but not the only, towards solving the problem.
More to the point, I intend to participate in this program. This is an area close to my heart (and brain) and I would definitely like to think and develop strong event based approaches. As I make progress in my thinking on some of these ideas, I will be sharing my ideas here with you in the hope that some of you will provide me interesting feedback on this.