[Colloquium] REMINDER: Talks at TTIC: Ayan Chakrabarti, Harvard University

Dawn Ellis dellis at ttic.edu
Wed Apr 2 09:34:31 CDT 2014


When:     Thursday, April 3rd at 11am

Where:    TTIC, 6045 S Kenwood Avenue, 5th Floor, Room #526

Speaker:  Ayan Chakrabarti, Harvard University

Title:       Spatio-Spectral Models for Computer Vision and Computational
              Photography

Abstract:

Inferring scene properties from images requires modeling and exploiting the
structure and statistical redundancies that exist in the visual world. In
this talk, I will discuss one aspect of this structure---that which exists
in the spatial variations of different spectral (i.e., color) channels in
local image regions. I will begin by describing an empirical study of these
statistics based on a new database of hyperspectral images of real-world
scenes. Then, I will describe how intuitions from this study led to the
development of two new computer vision systems, each featuring a modified
sensor/camera design and novel inference algorithm. The first system is a
single-shot camera that simultaneously produces both a color image and a
depth map by inducing a particular type of wavelength-dependent defocus
blur. The second system is a single-shot camera that produces a
conventional color image, but with a light efficiency that is nearly three
times greater than current color camera designs. The keys to the success of
both systems are inference algorithms that are co-designed with the optics,
and that exploit the spatio-spectral structure that exists in natural
images.

Bio:

Ayan Chakrabarti received the B. Tech and M. Tech degrees in electrical
engineering from the Indian Institute of Technology Madras, Chennai, India,
in 2006, and the S.M. and Ph.D. degrees in engineering sciences from
Harvard University, Cambridge, MA, in 2008 and 2011, respectively. He is
currently a post-doctoral fellow at Harvard University, where his research
focuses on the development of statistical models and the design of imaging
sensors and inference algorithms for applications in computer vision and
computational photography.

Host:  Greg Shakhnarovich, greg at ttic.edu


-- 
*Dawn Ellis*
Administrative Coordinator,
Bookkeeper
773-834-1757
dellis at ttic.edu

TTIC
6045 S. Kenwood Ave.
Chicago, IL. 60637
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.cs.uchicago.edu/pipermail/colloquium/attachments/20140402/acbe0ab3/attachment.htm 


More information about the Colloquium mailing list