Automatic video production of lectures using an intelligent and aware environment

ABSTRACT

This paper makes the case that much of the promise of ubiquitous multimedia depends on the availability of valued material. In business and academic environments the "presentation in a lecture room with laptop graphics" is a common way of communicating, but making a presentation readily available outside of the room is still a challenge because of the complexities of capturing and distributing the material. The AutoAuditorium™ System creates a multi-camera video program of a lecture in real time, without any human control beyond turning the system on and off. The system reduces the opportunity costs of making such a program to the point that it gets used for events previously not seen as candidates for video. Thus an event does not need nearly as many viewers to be considered worth capturing and many more events are seen by many more people.This paper presents a quick overview of the AutoAuditorium System technology and operational characteristics, a history of it's ancestry, development and use, and some usage experiences that demonstrate its current utility and future potential.The AutoAuditorium System is an example of an "intelligent and aware environment." In particular, it is:intelligent - about creating multi-camera television programs of lectures, in real time, with one or more people on a stage using projected visuals.aware - of the motion and gesturing of the people on stage.- of changes in the projected visuals.Originally created in the early- and mid-1990s as a research project [1] at Bellcore (Bell Communications Research, now Telcordia Technologies), it has been available as a commercial product from Foveal Systems since 1999.

References

1. Bianchi, M. H. AutoAuditorium: a fully automatic, multi-camera system to televise auditorium presentations. In Joint DARPA/NIST Smart Spaces Technology Workshop (1998). www.AutoAuditorium.com/nist/autoaud.html.
2. Bianchi, M. H. Autoauditorium system home page, 2004. www.AutoAuditorium.com.
3. Brotherton, J. A. Enriching Everyday Activities through the Automated Capture and Access of Live Experiences; eClass: Building, Observing and Understanding the Impact of Capture and Access in an Educational Domain. PhD thesis, Georgia Institute of Technology, July 2001.
4. Christianson, E. A. Declarative camera control for automatic cinematography. In Proceedings of the Thirteenth National Conference on Artificial Intelligence (1996), pp. 148--155.
5. E. Machnicki, L. R. Virtual director: Automating a webcast. In Proceedings of SPIE Vol. 4673: Multimedia Computing and Networking (2002), pp. 208--225.
6. Kernmani, P. The IBM research media portal, December 2002. www.poly.edu/Podium/eef2002.cfm.
7. Michael Gleicher, J. M. Towards virtual videography. In ACM Multimedia 2000, Los Angeles, CA (November 2000). www.cs.wisc.edu/graphics/Papers/Gleicher/Video/vv.pdf.
8. Rui, Y., He, L., Gupta, A., and Liu, Q. Building an intelligent camera management system. In Proceedings of ACM Multimedia 2001 (2001), pp. 2--11.
9. Steinmetz, A. A. E-seminar lecture recording and distribution system: a bottom up approach from video to knowledge streaming. In Proceedings of SPIE Volume 4312 (2001). www.spie.org/web/meetings/programs/pw01/confs/4312.html.
10. Truong, K., Abowd, G., and Brotherton, J. Who, what, when, where, how: Design issues of capture and access applications. In Ubicomp 2001 Conference Proceedings (2001), ACM Press, pp. 209--224.
11. Yuanchun Shi, E. A. The smart classroom: Merging technologies for seamless tele-education. In IEEE Pervasive Computing Magazine (April-June 2003).

Index Terms

• Applied computing

  • Education

• Computing methodologies

  • Artificial intelligence

    • Computer vision

      • Computer vision tasks

  • Machine learning