For the first 15 minutes, we begin with Light Breakfast. Registration and sign-in. Walk-ins are welcome.
From 10:15--11:45,  we will use MATLAB and the Computer Vision System Toolbox to demonstrate applications of object detection and tracking. Through interactive examples, you will see how to:
•             Recognize objects using scale- and rotation-invariant features
•             Use a cascade-classifier algorithm to find objects of interest
•             Track objects using a point tracking algorithm
•             Perform Kalman Filtering to predict the location of a moving object
•             Implement a motion-based multiple object tracking system
 This event assumes some experience with MATLAB and Image Processing Toolbox.
From 11:45--12, Q&As with the speaker and Allyson Butler, Education Department, MathWorks.
 
Please register at the speaker's link.
 

In their attempt to mimic the proof of Fermat's Last Theorem for GL(n), Clozel, Harris, and Taylor, were led to a conjectural analogue of Ihara's lemma -- which is still open for n>2. In this talk we will revisit their conjecture from a more modern point of view, and reformulate it in terms of local Langlands in families, as currently being developed by Emerton and Helm. At the end, we hope to hint at potential applications.

Polynomial interpolation is well understood on the real line. In
multi-dimensional spaces, one often adopts a well established one-dimensional method
and fills up the space using certain tensor product rule. Examples
like this include full tensor construction and sparse grids construction.
This approach typically results in
fast growth of the total number of interpolation nodes and certain fixed geometrical structure of the
nodal sets. This imposes difficulties for practical applications,
where obtaining function values at a large number of nodes is
infeasible. Also, one often has function data from nodal locations that are not
by "mathematical design" and are ``unstructured''.

In this talk, we present a mathematical framework for conducting polynomial interpolation
in multiple dimensions using arbitrary set of unstructured nodes. The resulting method,
least orthogonal interpolation, is rigorous and has a straightforward numerical implementation.
It can faithfully interpolate any function data on any nodal sets, even on those that are considered
degenerate by the traditional methods. We also present a strategy to choose
``optimal'' nodes that result in robust
interpolation. The strategy is based on optimization of Lebesgue
function and has certain highly desirable mathematical properties.

We consider discrete quasi-periodic long range operators with Liouvillean frequency. First, based on generalized Gordon type argument, we show that they can be approximated by a sequence of finite range operators which have no point spectrum for any phase. On the other hand, we show that when the potential for the dual model is small, then they can be approximated by a sequence of long range operators which have at least one eigenvalue for each phase in a set of full measure.

We prove that the resolvent set of any (possibly singular)
almost periodic Jacobi operator is characterized as the set of all
energies whose associated Jacobi cocycles induce a dominated splitting.
This extends a well-known result by Johnson for Schrödinger operators.