Colloquium Series 2014-2015
Here is a list of our speakers for the 2014 and 2015 academic year:
September 30, 2014 - Chris Honey (University of Toronto)
November 25, 2014 - Ben Thompson (University of Waterloo)
December 16, 2014 - Graham Taylor (University of Guelph)
February 24, 2015 - Melvyn A. Goodale (University of Western Ontario)
April 8, 2015 - Waterloo Brain Day (9th Annual)
Date:
September
30,
2014
Location:
PAS
2464
Time:
3:30
p.m.
Speaker:
Chris
Honey
(University
of
Toronto)
Title:
Uncovering
Stimulus-Induced
Network
Dynamics
during
Narrative
Comprehension
Real
world
cognition
requires
the
coordination
of
information
processing
between
modalities
(e.g.
auditory
and
visual)
and
systems
(e.g.
language
and
memory).
Therefore,
the
brain
must
continually
reorganize
interactions
between
and
within
networks.
I
will
describe
a
new
method
for
mapping
these
network
changes,
and
how
it
can
reveal
precise
and
reliable
network
dynamics
during
the
perception
of
a
naturalistic
narrative
stimulus.
Date:
November
25,
2014
Location:
PAS
2464
Time:
3:30
p.m.
Speaker:
Ben
Thompson
(University
of
Waterloo)
Title:
Learning
to
See
with
a
“Lazy
Eye”;
Harnessing
Visual
Cortex
Plasticity
to
Treat
Amblyopia
Amblyopia is a neurodevelopmental disorder of the visual cortex that is often considered to be untreatable in adulthood due to insufficient neural plasticity. I will present a series of studies which indicate that both binocular perceptual learning and non-invasive brain stimulation techniques can improve visual function in adult patients with amblyopia, possibly by reducing inhibitory interactions within the visual cortex. I will also present new data from a recent study on the use of the anti-depressant drug citalopram to promote visual cortex plasticity in adults with amblyopia.
Date:
December
16,
2014
Location:
PAS
2464
Time:
3:30
p.m.
Speaker:
Graham
Taylor
(University
of
Guelph)
Title:
Learning
Representations
with
Multiplicative
Interactions
Representation learning algorithms are machine learning algorithms which involve the learning of features or explanatory factors. Deep learning techniques, which employ several layers of representation learning, have achieved much recent success in machine learning benchmarks and competitions, however, most of these successes have been achieved with purely supervised learning methods and have relied on large amounts of labeled data. In this talk, I will discuss a lesser-known but important class of representation learning algorithms that are capable of learning higher-order features from data. The main idea is to learn relations between pixel intensities rather than the pixel intensities themselves by structuring the model as a tri-partite graph which connects hidden units to pairs of images. If the images are different, the hidden units learn how the images transform. If the images are the same, the hidden units encode within-image pixel covariances. Learning such higher-order features can yield improved results on recognition and generative tasks. I will discuss recent work on applying these methods to structured prediction problems.
Date: February 24, 2015
Location:
PAS
2464
Time:
3:30
p.m.
Speaker:
Melvyn
A.
Goodale
(University
of
Western
Ontario)
Title:
How
We
See
and
Hear
Stuff:
Visual
and
Auditory
Routes
to
Understanding
the
Material
Properties
of
Objects
Almost all studies of object recognition, particularly in brain imaging, have focused on the geometric structure of objects (i.e. ‘things’). Until recently, little attention has been paid to the recognition of the materials from which objects are made (i.e. ‘stuff’), information that is often signalled by surface-based visual cues (the sheen of polished metal) as well as auditory cues (the sound of water being poured into a glass). But knowledge about stuff (the material properties of objects) has profound implications, not only for understanding what an object is, but also for the planning of actions, such as the setting of initial grip and load forces during grasping. In recent years, our lab has made some headway in delineating the neural systems that mediate the recognition of stuff (as opposed to things), not only in sighted people but also in blind individuals who use echoes from tongue clicks to recognize the material properties of objects they encounter. I will discuss evidence from both neuropsychological and fMRI studies demonstrating that lateral occipital regions in the ventral stream play a critical role in processing the 3-D structure and geometry of objects, whereas more anteromedial regions (particularly areas in the parahippocampal gyrus and collateral sulcus) are engaged in processing visual and auditory cues that signal the material properties of objects.