[Colloquium] Computation of Imaging-Based Science: Yali Amit, Richard Kron - TODAY!

[Ninfa Mayorga]

Computation Institute: Computation of Imaging-Based Science

Speakers: Yali Amit (Statistics) and Richard G. Kron (Astronomy and Astrophysics)
Host: Gordon L. Kindlmann 
Date: Nov 30, 2011
Time: 2:30 PM - 5:30 PM
Location: University of Chicago, Searle 240A, 5735 S. Ellis Avenue

Computation of Imaging-Based Science

All visitors are welcome to join the next meeting of the Computation of Imaging-Based Science seminar, with two speakers at 2:30 and 4:15. Those not able to attend in person are welcome to watch and listen in via Adobe Connect, at http://anl.adobeconnect.com/cibs/

2:30: Multiple object models in computer vision.  The goal of Computer Vision is the automatic labeling of images containing multiple objects as well as noise and clutter. Recent work has focused on two main tasks. The first is the classification among object classes in segmented images containing only one object and the second is the detection of a particular object class in a large image. Both tasks have been primarily addressed using discriminative learning. It is not clear however how these methods can extend to deal with the recognition of multiple object classes in images containing a number of objects in a wide range of configurations. I will present an approach which starts from simple statistical models for individual objects. With these models the important notion of invariance can be clearly formulated. Furthermore the individual object models can be composed to define models for object configurations. Decisions are likelihood based and do not depend on pretrained decision boundaries. The model formulation also leads to a coarse to fine strategy for efficient computation of the optimal scene annotation. These ideas will be illustrated in several applications reading handwritten zipcodes, detecting faces, and tracking vesicles in video microscopy.

4:15: I will review the Dark Energy Survey: its scientific goals, and how we intend to use the image data from the Dark Energy Camera to address those goals. "Dark Energy" is a term used to describe the surprising acceleration of the expansion rate of the Universe. At present, there is no generally accepted theoretical basis for this phenomenon. However, by characterizing the acceleration with greater precision (how much acceleration and whether it changes with time), we can start to constrain some classes of models. There are classically four complementary approaches to measuring the expansion rate of the Universe, and the DES will use all of them. 1) Measure the brightness of supernovae as an indicator of distance. 2) measure the angular size of "baryon acoustic oscillations" as an indicator of distance. 3) measure the number of massive clusters of galaxies per unit volume as an indicator of the growth of structures under gravity. 4) measure the mass surface density as a function of distance by measuring the (very small) distortion of distant galaxies by foreground mass; this is another way to measure the gravitational growth of structures.

Further information about these and future speakers is available here:
http://www.ci.uchicago.edu/wiki/bin/view/CIBS


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.cs.uchicago.edu/pipermail/colloquium/attachments/20111130/f74b1d3d/attachment.htm