Michael A. Smith Ph.D.
Department of Electrical and Computer Engineering Class of 1998
Carnegie Mellon University
Pittsburgh, PA 15213
msmith (AT) cs.cmu.edu


Automated Video Skimming


Research


Publications

Evolving Video Skims into Useful Multimedia Abstractions
Christel, M.G., Smith, M.A., Taylor, C.R, & Winkler, D.B. Proceedings of the CHI '98 Conference on Human Factors in Computing Systems. (Los Angeles, CA, April, 1998).

Video Skimming for Quick Browsing Based on Audio and Image Characterization,
Smith, M., Kanade, T., Carnegie Mellon technical report CMU-CS-97-111, February 1997. To appear in CVPR97. Low resolution postscript version ( low res. version , 1.2 Meg.)

Abstract and [MOVIE] Sample Skim
16768 frames (559 sec) reduced to 1666 frames (55.5 sec), 10:1 compaction

Intelligent Access to Digital Video: The Informedia Project,
Wactlar, H., Kanade, T., Smith, M., Stevens, S.Smith, M., IEEE Computer, Vol. 29, No. 5, May 1996
Automating the Creation of a Digital Video Library,
Michael A. Smith and Michael G. Christel, ACM Multimedia, 1995

Text, Speech, and Vision for Video Segmentation: The Informedia Project,
Michael A. Smith and Alexander Hauptmann, AAAI Fall 1995 Symposium on Computational Models for Integrating Language and Vision Full resolution postscript version.

Video Segmentation in the Informedia Project,
Hauptmann, A.G. and Smith, M, IJCAI-95, Workshop on Intelligent Multimedia Information Retrieval. Montreal, Quebec Canada. August 1995