[4eyes] Reading Group tomorrow

[Chris Sweeney]

We will have a reading group tomorrow at 11am in the MAT conference room
before the lab meeting. Sorry for the delay... we in the Foglab just
settled on a paper about 5 minutes ago

The paper is called RECON (Residual Consensus) and builds on the idea of
RANSAC. The primary idea is roughly that models generated from enough
inliers (aka a "good" model) will have similar behavior of their residuals,
whereas bad models will have incoherent residual behavior. Using this, the
authors discover a way to combine good models by examining the residuals.

http://www.cs.unc.edu/~rraguram/papers/0727.pdf

Don't stress about knowing the paper really well before tomorrow, just come
tomorrow with the paper printed out and be ready to engage in the
discussion!

-Chris
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cs.ucsb.edu/pipermail/ilab-users/attachments/20130131/c7eb433b/attachment.html>