UTCS Colloquia /AI - Rob Fergus/New York University, "Deconvolutional Networks", JGB 2.218

Contact Name: 
Jenna Whitney
Date: 
Apr 15, 2011 2:00pm - 3:00pm

There is a sign-up schedule for this event that can be found at

http://www.cs.utexas.edu/department/webevent/utcs/events/cgi/list_event

s.cgi

Type of Talk: UTCS Colloquia /AI

Speaker/Affiliation: Rob F

ergus/New York University

Talk Audience: UTCS Faculty, Grad Students

, Undergrads, and Outside Interested Parties

Date/Time: Friday, Apri

l 15, 2011, 2:00 p.m.

Location: JGB 2.218

Host: Kristen Grauman

Talk Title: Deconvolutional Networks

Talk Abstract:
We present
a hierarchical model that learns image decompositions via alternating laye

rs of convolutional sparse coding and max pooling. When trained on natural

images, the layers of our model capture image information in a variety of

forms: low-level edges, mid-level edge junctions, high-level object parts
and complete objects. To build our model we rely on a novel inference sche

me that ensures each layer reconstructs the input, rather than just the ou

tput of the layer directly beneath, as is common with existing hierarchica

l approaches. This makes it possible to learn multiple layers of representa

tion and we show models with 4 layers, trained on images from the Caltech-

101 and 256 datasets. Features extracted from these models, in combination
with a standard classi