Skip to main content.
Advanced search >
<< Back to previous page Print

<< Wednesday, May 01, 2013 >>


Remind me

Tell a friend

Add to my Google calendar (bCal)

Download to my calendar

Bookmark and ShareShare


[Dissertation Talk] Shapes, Paint, and Light

Seminar: Departmental | May 1 | 12-1 p.m. | Sutardja Dai Hall, 254 SDH


Jonathan Barron, UC Berkeley

Electrical Engineering and Computer Sciences (EECS)


A core problem in computer vision is that of inferring the intrinsic, 3D structure of the world from flat, 2D images of that world. Traditional methods for recovering scene properties such as shape, reflectance, or illumination rely on multiple observations of the same scene to over-constrain the problem. Recovering these same properties from a single image seems almost impossible in comparison --- the space of shapes, paint, and lights that exactly reproduce a single image is vast.

However, certain explanations are more likely than others: surfaces tend to be smooth, paint tends to be uniform, and illumination tends to be natural. We therefore pose this problem as one of statistical inference, and define an optimization problem that searches for the *most likely* explanation of a single image. Our resulting technique can be viewed as a superset of several classic computer vision problems, and outperform all previous solutions to those constituent problems.


jonbarron@gmail.com, 857-231-2248