Reasoning about a scene from a photograph is an inherently ambiguous task. This is because a single image in itself does not carry enough information to disambiguate the world that it is depicting. Of course, humans have no problems understanding photographs because of all the prior visual experience they can bring to bear on the task. How can we help computers do the same? Our solution is to use large amounts of visual data, both labeled and unlabeled, as a way of capturing the statistics of the natural world.

In this talk, I will present some of our recent results on inferring geometric, photometric, and geographic scene properties from a single image. I will first briefly describe our system for estimating the rough geometric surface layout of a scene. I will show how this information, in turn, can be useful for modeling objects in the scene. Next, I will describe an approach for using the surface layout information as a way of estimating a rough illumination map for the scene. Finally, I will describe a new system that uses millions of unlabeled photographs from Flickr to capture some implicit geographic scene structure of an image.

Alexei (Alyosha) Efros is an assistant professor at the Robotics Institute and the Computer Science Department at Carnegie Mellon University. His research is in the area of computer vision and computer graphics, especially at the intersection of the two. He is particularly interested in using data-driven techniques to tackle problems which are very hard to model parametrically but where large quantities of data are readily available. Alyosha received his Ph.D. in 2003 from the University of California, Berkeley and spent the following year as a fine fellow at Oxford, England. Alyosha is a recipient of the NSF CAREER award (2006), the Sloan Fellowship (2008), and the Guggenheim Fellowship (2008).