Model-Based Statistics for Analyzing Images and Videos
Deva Ramanan, University of California, Irvine

The digital media revolution is readily underway; staggering amounts of images and videos are available in online repositories. The amount of visual data is too much to process by hand; one of the challenges that lies ahead is the design of automated systems for processing such data. However, everyday scenes from our three-dimensional visual world are complex and can be difficult to explicitly model in all their detail. One attractive approach is to leverage statistical tools that take advantage of "big data" to encode such complex phenomena. I will discuss one such approach that integrates statistical models with knowledge about three-dimensional geometry.

Background review article

Ramanan, Deva. "Part-based models for finding people and estimating their pose." Visual Analysis of Humans. Springer London, 2011. 199-223.

Loading more stuff…

Hmm…it looks like things are taking a while to load. Try again?

Loading videos…