Oxyopias Schedule
Wednesday, December 8th, noon in room 489 Minor Hall
Ian G. Cox
Research Fellow at Research Clinic Bausch & Lomb, Rochester, NY
Host: Ken Polse
The effect of lens material and design features on lens movement - in-vitro and in-vivo investigations
Abstract
The role of lens material and design features in on-eye lens movement
has historically been driven by clinical observation using commercially
available lens products. Unfortunately this invariably leads to a number of
uncontrolled lens material and design variables that confound the clinical
results and lead to erroneous hypotheses. This lecture describes a series of
controlled clinical studies to determine the effect of lens material and design
features on lens movement and centration. Attempts to confirm these findings
with laboratory off-eye models, and theoretical models will also be covered.
December 3 , 11:am - 12:30 pm 2515 Tolman
Jitendra Malik
Computer Science, University of California at Berkeley
Co-sponsored by Institute for Cognitive Studies
Host: Steve Palmer
Computational Mechanisms Of Visual Grouping
Abstract
I shall argue that early and intermediate level processing in the visual
system be modeled as a three stage process. The first stage is a
measurement stage carried out in V1 with spatiotemporal receptive fields
tuned to orientation, spatial frequency, opponent color, and short-range
motion. The second stage is a grouping stage resulting in the formation
of regions of coherent brightness, color and texture. Call these
'proto-surfaces'. The third stage results in the formation of
surfaces/objects with attached properties such as lightness, object
motion, occlusion relationships (figure-ground), depth, slant-tilt etc
and is based on the combined operation of Gestalt grouping factors,
shape cues, and can be partially influenced by knowledge of familiar
configurations.
The first stage can and has been modeled by many researchers using tools
of linear system analysis. We offer a novel approach to the second stage
by modeling it as the process of finding a partition of the image into
regions such that there is high similarity within a region and low
similarity across regions. This is made precise as the 'Normalized cut'
criterion which can be optimized by solving an eignevalue problem. The
resulting eigenvectors provide a herarchical partitioning of the image
into regions ordered according to salience. Brightness, color, texture,
motion similarity, proximity and good continuation can all be encoded
into this framework. We show results on complex images of natural scenes
which demonstrate the significant superiority of this technique over
classical approaches based on edge detection. Phenomena such as
subjective contours emerge as side consequences.
Our work on the third stage is preliminary; I shall argue on
computational and psychophysical grounds that modular shape processing
should be abandoned, and that grouping driven by ecological statistics
is as crucial as shape cues driven by ecological optics.
This talk is based on joint work with Jianbo Shi, Thomas Leung and Serge
Belongie.
October 29, Noon, 489 Minor Hall
Laurent Itti
California Institute of Technology - Computation and Neural Systems
Pasadena, CA 91125
Host: Lawrence Stark
A Bottom-Up Search Mechanism for Over and Covert Shifts of Visual Attention
Abstract
Most models of visual search, whether involving overt eye movements or
covert shifts of attention, are based on the concept of a saliency map,
that is, an explicit two-dimensional map that encodes the saliency or
conspicuity of objects in the visual environment. Competition among neurons
in this map gives rise to a single winning location that corresponds to the
next attended target. Inhibiting this location automatically allows the
system to attend to the next most salient location. We describe a detailed
computer implementation of such a scheme, focusing on the problem of
combining information across modalities, here orientation, intensity and
color information, in a purely stimulus-driven manner. The model is
applied to common psychophysical stimuli as well as to a very demanding
visual search task. Its successful performance is used to address the
extent to which the primate visual system carries out visual search via one
or more such saliency maps and
how this can be tested.
October 22 Friday, Noon 489 Minor Hall
Michael Landy
Dept of Psychology, New York Univeristy
Host: Marty Banks
Exploring Early Vision Using Oriented Textures
Abstract
Substantial evidence indicates that early visual coding includes
both 1st-order and 2nd-order mechanisms. "1st-order mechanisms" is
just another name for the now-familiar, linear, spatial-frequency
channels. 2nd-order mechanisms involve a 2nd-order linear filter
that follows an intervening nonlinearity. We have used oriented
textures (either randomly placed oriented line segments or filtered
noise) to investigate these 2nd-order spatial mechanisms. Results
indicate that:
-
there is a preference for 1st- and 2nd-order filters to be orientationally aligned;
-
some texture pairs are discriminable for which there are no built-in texture edge mechanisms;
-
there are multiple, bandpass, spatial-frequency tuned, 2nd-order filters just as in the 1st-order case.
The third conclusions stems from a continuing series of experiments that
applies the techniques used to characterize 1st-order filters (e.g.,
contrast sensitivity, summation, increment threshold, etc.) to the
2nd-order filters.
October 8 Friday, 2:00 Noon, 489 Minor Hall
Jay Enoch
School of Optometry, Group in Vision Science, UC Berkeley
Host: Cliff Schor
The first known lenses: Remarkable Egyptian lenses from ca. 4600 year
Abstract
They were parts of apparent schematic eyes in Egyptian statues. Included was
a lively visual illusion where the observer seemed to be followed by the
pupils of these eyes as he/she rotated about the statue. The illusion will
be demonstrated in an accompanying model.
October 1 Friday ,12:00 noon, 489 Minor Hall
Maureen K. Powers
Visiting professor in Molecular and Cell Biology, UC Berkeley
Host: Marty Banks
What can a regenerated retina see?
Abstract
The central nervous systems of some fish and amphibians continue to add new
neurons in adulthood, unlike humans, and they have the remarkable capacity
to regenerate following traumatic injury, also unlike humans. The source
of the normal and injury-induced neuronal addition in fish retina is a
special class of stem-like cell which normally divides throughout life.
Functional studies have been aimed at determining whether the new retinal
cells that are formed in turn form new projections to the brain, and
whether the new retina-brain complex can support vision. We find that
connections to the optic tectum do return, and that the electroretinogram,
optokinetic nystagmus, dorsal light response, and the ability to respond to
dim lights in a classical conditioning paradigm all return as well during
regeneration. The quality of vision seems generally to be correlated with
the quality of retinal regeneration as assessed morphologically. However,
even in the best cases visual capacity is diminished: Animals are less
sensitive scotopically, have poor spatial vision and are probably deficient
in color vision. Retinal disorganization, non-optimal connections and
abnormal photopigment complements may be responsible for the deficits.
Although imperfect, the natural retinal regeneration process in goldfish
clearly does support sensory function, leading to the hope that retinas in
other vertebrates (ultimately humans) might be induced to do the same.
September 29, Wednesday 12:00 noon, 489 Minor Hall
Bart Anderson
MIT, Dept of Brain & Cognitive Sci, Cambridge MA
Host: Cliff Schor
Synthesizing contours from moving images
Abstract
One of the fundamental problems of recovering scene geometry involves the
computation of occlusion relationships. Both stereopsis and motion
generate spatio-temporally unmatched image regions along occlusion
boundaries, yet both domains have been dominated by computational
strategies that require corresponding features to be present in the
multiple views. I will describe a new class of contour synthesis
mechanisms that we have discovered in the motion domain that relies on the
presence of these unmatched features, a model of contour synthesis, and the
relevance of these findings for statistical (Bayesian) models of scene
interpretation.
September 24, Noon ,489 Minor Hall
Pablo Artal
Laboratorio de Optica,, Universidad de Murcia, Spain
Host: Stan Klein
Aberration Measurement And Correction In The Human Eye
Abstract
A real-time wavefront sensor for the human eye will be described. New
results on the aberrations' temporal dynamics and on the relative
contribution of the ocular components (cornea and lens) to the overall
eye's aberrations will be presented. Strategies for correcting the ocular
aberrations and their impact in vision and high resolution retinal imaging
will be also discussed.
September 17 ,Noon , 489 Minor Hall
Lynn Marran PhD, OD
UC Berkeley, School of Opometry
Host: Cliff Schor
Aniso-Accommodation and its clinical implications?
Abstract
Aniso-accommodation is unequal accommodation of the eyes occurring
simultaneously while in the binocular state. Marran and Schor(1998)
demonstrated that 0.75D of aniso-accommodation could be reliably elicited in
subjects. Previous (Ogle, 1937; Spencer et al 1954; Stoddard & Morgan,
1942; Ball, 1952) and recent (Flitcroft, 1992; Koh et al, 1998; Carlin et
al, 1998;) investigators have had equivocal results in demonstrating
aniso-accommodation. This talk will address the unique characteristics of
the response and hence the conditions that are required to evoke and measure
it. These characteristics both explain the equivocal results and also
suggest what role aniso-accommodation may play in the non-laboratory
setting. In particular, a possible role of aniso-accommodation in the
induction / or recovery from anisometropia will be discussed.
August 27 Friday, 12:00 Noon, 489 Minor Hall
Mitesh Kapadia
Univ. Pennsylvania Neuroscience and Rockefeller University
Lateral interactions in primary visual cortex and their role in perception
July 27 Tuesday, 12:00 Noon, 489 Minor Hall
Jan Koenderink
Utrecht Biophysics, Fysisch Laboratory RUU, Netherlands
Host: Jitendra Malik
Pictorial Relief
Abstract
When the human observer looks at a photograph of a scene the perception
is of a curious twofold nature: One sees the photograph as an object (a
flat piece of paper with various pigmentation in a certain simultaneous
order) and at the same time one looks into the photograph into
"pictorial space." Pictorial space is three, rather than two
dimensional. When the pictorial space is like the physical space of the
scene at the time the photograph was taken one calls that perception
veridical. Since photographs typically underdetermine the scene one
expects such veridical perceptions to be rare and due to either change
occurrance or prior knowledge. When it is known which major cues are
used by the observer, it is possible to describe the remaining ambiguity
formally. In general bilinear problems (for instance "shape from
shading") this ambiguity is a shear in threespace that leaves the image
plane invariant. I will show examples of nonveridical perceptions that
differ by exactly these transformations but are metrically different to
a surprising degree. Human observers appear to apply such ambiguity
transformations in order to "move the mind's eye" in pictorial space.
June 25, 12:00 Noon, 489 Minor Hall
Gopathy Purushothaman
University of Houston
Host: Ralph Freeman
Oscillatory visual percepts and their putative neural mechanisms
Abstract
It has been known for over a century that visual percepts of very bright
static flashes and moving objects have a repetitive nature, and have been
referred to as Recurrent Images and "Charpentier's Bands,"
respectively (Bidwell, 1894; Charpentier, 1896; McDougall, 1904). Early
explanations of these phenomena attributed these percepts to the rhythmic
discharges observed in some retinogeniculocortical neurons (Frohlich, 1912;
Buttneret al, 1969). The recent physiological investigations of
theseoscillatory activities and their putative functional roles motivateda
re-examination of these psychophysical phenomena. In a series of studies,
we measured psychophysically.
-
the impulse responses to bright stationary flashes,
-
the perceived number of recurrences in the percept of pairs of flashes, and
-
the Charpentier bands in the smear of bright moving targets.
Our data show significant inter-subject, inter-trial and intra-trial
variabilities in several aspects of the oscillations, including their frequencies
and latencies. Despite these variabilities, a computational model can be used to
show that the temporal impulse response to a stationary flash can account for the
perceived number of recurrences in static flashes and the patterns of the
Charpentier bands.
This study suggests that neural oscillations are unlikely to constitute a robust
internal code. If oscillations and/or their coordination were an ideal code, then
they would be inaccessible at the perceptual level. Our study, on the other hand,
shows that, under certain stimulus conditions, the putative code (oscillations)
can interfere with the information that it carries (perceived quality of the
stimulus).
May 21, 12:00 Noon, 489 Minor Hall
Andrew Parker
University Lab of Physiology, Oxford
Host: Marty Banks
Binocular Neurons and the Perception of Depth
Abstract
The first stage in the primary visual pathways at which there is
significant binocular convergence lies in cortical area V1. Many neurons
in V1 are highly sensitive to binocular disparity and they undoubtedly play
an important intermediate role in the detection of stereoscopic depth. I
will present several results from our laboratory that highlight
discrepancies between the psychophysical perception of stereoscopic depth
and the characteristics of these V1 neurons. These results point to an
important role for extrastriate cortex in stereoscopic vision.
May 3 (Monday), 12:00 Noon, 489 Minor Hall
Nansi Jo Colley
Ophth and Vis Sci and Genetics, University of Wisconsin, Madison, WI
Host: Sheldon Miller
Molecular Genetics of the Drosophila Visual System
Abstract
The overall objective of our research program is to utilize Drosophila as a model
for studying hereditary human diseases that cause retinal degeneration and
eventual blindness. The complexity and variations of retinal degeneration
disorders has lead to the identification of multiple subtypes, each with a
distinct genetic and biochemical basis. This complexity, the infrequent
availability of ocular tissues from patients, and the broad base of knowledge of
Drosophila genetics combine to make Drosophila a powerful animal model for
studying human disease. We are engaged in two areas of research. One set of
projects is focused on rhodopsin trafficking in the secretory pathway and the
other deals with the role of calcium in photoreceptor function. To understand
mechanisms of calcium modulation, we have initiated the characterization of a
novel sodium/calcium-potassium exchanger in Drosophila, NCKX30C.
Photoreceptors in both Drosophila and humans alike, utilize a photopigment,
rhodopsin, for vision. During biosynthesis, rhodopsin is specifically targeted
to the site of vision within the photoreceptors. One goal of our studies is to
further elucidate the mechanisms of protein targeting and trafficking in the
photoreceptor cells. We have previously shown that defects in rhodopsin
trafficking result in severe retinal defects and retinal degeneration in
Drosophila. Similar mechanisms are thought to occur in some human retinal
degeneration disorders. In fact, more than twenty five percent of all autosomal
dominant retinitis pigmentosa (ADRP) cases appear to be caused by mutations in
the gene encoding the visual pigment rhodopsin, located on human chromosome 3q.
Over sixty distinct mutations in rhodopsin have now been reported in ADRP
patients. The mechanism by which the mutant rhodopsin proteins cause dominant
retinal degeneration has been actively investigated in Drosophila as well as in
vertebrates. In screening mutagenized Drosophila chromosomes we have recovered
thirteen mutations in the gene encoding rhodopsin. All 13 mutations entail
single nucleotide changes, which together define 10 different sites in the
protein. Remarkably, four such missense mutations in Drosophila (G119E, P184L,
E194K and G195S) correspond to mutations in exactly the same amino acid residues
found in known human ADRP (G106R, P171L, E181K and G182S, respectively). We have
demonstrated that in these cases, retinal degeneration results from interference
of the assembly of mature wild type rhodopsin by the mutant proteins.
Intracellular free calcium influences several distinct time-dependent cellular
events in human and Drosophila photoreceptors. Therefore, the precise control of
spatial and temporal profiles of calcium are extremely important for
photoreceptor function. In addition, the prolonged elevation of cytosolic
calcium can be toxic and could lead to retinal degeneration. Cells maintain
their low cytosolic calcium levels using ATP-driven pumps and
sodium/calcium(-potassium) exchangers. To understand the molecular basis of
calcium modulation, we are characterizing a sodium/calcium-potassium exchanger
(NCKX30C) that is expressed in the Drosophila photoreceptors. Genetic screens
for mutations in the Drosophila visual system continue to yield exciting
phenotypes that define novel photoreceptor genes and mechanisms for
photoreceptor function and retinal degeneration in vivo.
April 29 (Thursday), 3PM, 489 Minor Hall
Fiona Stapleton
Host: Suzi Fleiszig
Epidemiology, basic and clinical science approaches in understanding contact lens related disease
Abstract
Epidemiological studies of ocular disease provide information
on their frequency, distribution and associated risk factors. This type of data
can assist in the management of disease but also can initiate basic and clinical
science approaches to testing specific hypotheses relating to the pathogenesis of
disease.
This methodology has been used in the evaluation of the pathogenesis of contact
lens related disease. Initial anecdotal reporting of lens related disease during
the 1980's led to the development of large scale prospective cohort and case
control studies to estimate the size of the problem and to determine specific
risk factors. In certain groups of complications, risk factors such as overnight
wear and the type of care system and its frequency of use were identified as
associated risk factors.
The identification of overnight lens wear as a major risk factor in certain
groups of complications led to the formulation of hypotheses of the impact of eye
closure coupled with lens wear on ocular defense mechanisms. Consequently,
increased levels of tear and serum proteins, cytokine and chemokine synthesis,
complement activation, leukocyte recruitment to the ocular surface and an
increase in ocular biota have all been demonstrated during uncomplicated eye
closure. Lens wear adds a further dimension to ocular defense during eye
closure, by affecting tear and membrane bound glycoproteins, altering leukocyte
recruitment, inhibiting tear exchange, providing a vector for bacterial carriage
and by providing a potentially antigenic stimulus. Hypothesis testing for each of
these different factors will be discussed
Ocular surface, contact lens and bacterial interactions are complex.
Understanding contact lens related complications clearly requires a
multidisciplinary approach to elucidate the pathogenesis and suggest preventive
strategies.
April 26 (Monday), 12:00 Noon, 489 Minor Hall
Lawrence O'Keefe
Center for Neuroscience, New York Univeristy
Host: Ralph Freeman
Motion Processing Mechanisms in the Primate Visual Cortex
Abstract
Our understanding of motion processing in the visual system is largely based on
studies using stimuli whose motion is cued by luminance (first-order motion).
However, there are many kinds of stimuli whose motion is cued not by luminance,
but rather by contrast, color, or texture ("second-order" motion).
Psychophysical studies suggest that first- and second-order motion information
may be processed by independent mechanisms. I wish to know how this
psychophysical dissociation is reflected in neural pathways -- is there is a
single neural pathway mediating motion perception, or do several neural
pathways each mediate a different aspect of motion perception?
My approach is to compare responses evoked by conventional first-order motion
targets to those evoked by a class of second-order motion targets designed to
elude detection by conventional motion mechanisms. I have used this approach in
combination with psychophysical techniques in human observers, single unit
recordings in areas V1 and MT of anesthetized monkeys, and more recently, single
unit recordings in area MT of alert monkeys. By comparing behavioral and
neuronal sensitivity to first- and second-order motion, I hope to identify and
characterize neuronal populations that mediate the perception of different kinds
of motion.
April 22 (Thursday), 3PM, 489 Minor Hall
Izumi Ozhawa
School of Optometry, University of
California, Berkeley
Functional Organization of the Primary Visual Cortex and Beyond
Abstract
Hierarchical organization and parallel processing in central visual pathways are
well established. However, the key roles played by specific cell types and the
details of neural circuitry are still not understood. Our exploration of general
receptive field characteristics of visual neurons provides clues about how
cortical neurons are built and what information they carry. I will discuss my
plans for investigating cortical neurons beyond the initial stages.
April 19 (Monday), 12:00 Noon, 489 Minor Hall
Andrew Goldberg
HHMI/Department of Biochemistry, University of Washington, Seattle, WA
Host: John Flannery
Towards an understanding of inherited retinal degenerations:
disease-linked mutations in peripherin/rds and rom-1
Abstract
Towards an understanding of inherited retinal degenerations: disease-linked
mutations in peripherin/rds and rom-1. The identification of heritable mutations
that cause progressive retinal dystrophies holds promise for understanding and
eventually treating many debilitating eye diseases. Inherited defects in the
photoreceptor-specific peripherin/rds and rom-1 polypeptides result in a
surprising and unprecedented heterogeneity of human retinal diseases. The
variety of clinical presentations underscores the importance of peripherin/rds
and rom-1 for human vision, but fails to reveal the molecular details of
pathogenesis. Using purified retinal photoreceptor membranes, I have shown that
peripherin/rds and rom-1 form the subunits of an integral membrane protein
complex that is restricted exclusively to the rim regions of both rod and cone
photoreceptor outer segment disks. This complex is proposed to play a role in
outer segment disk morphogenesis and stability. I have developed a cell culture
system for the heterologous expression and characterization of peripherin/rds and
rom-1, and used biochemical, immunochemical and biophysical analysis to probe
aspects of protein structure, function and role in human retinal
diseases.
April 15 (Thursday), 3PM, 489 Minor Hall
Michael Berry II
Department of Molecular and Cellular Biology, Harvard University
Host: Ralph Freeman
Anticipation of Moving Stimuli the Retina
Abstract
The visual system has a delay of ~50 ms in processing flashes of light, largely
arising in the retina. Such a delay poses severe problems for timing motor output
with moving objects (for instance, hitting a tennis ball). Using a flat array of
61 electrodes to record action potentials from up to 80 ganglion cells at once,
we have found that a moving bar elicits a traveling wave of ganglion cell
activity that peaks at or even ahead of the bar's leading edge. This
anticipation of moving objects mirrors a recently reported visual illusion:
observers perceive a moving bar to be ahead of a flash at the same location. The
measured spatial extent and temporal dynamics of a ganglion cell receptive field
cannot account for motion anticipation. However, a model that includes a contrast
gain control mechanism is highly successful. In some sense, anticipation is a
modest form of prediction that the eye makes about the near future.
April 8 (Thursday), 3PM 489 Minor Hall
Frank Schaeffel
University Eye Hospital Tuebingen, Experimental Ophthalmology
Hosts: Ian Bailey & Stan Klein
The visual control of eye growth: systems analysis, optics and biochemistry
Abstract
The length of the eye is normally properly matched to the focal length of cornea
and crystalline lens. However, in the industrial nations, there is a trend of
eyes becoming too long and myopic. Both experiments in animal models and studies
in humans suggest that visual experience has a significant impact on ocular
elongation. To prevent exaggerated growth and myopia, two approaches emerge: (1)
restriction of the critical visual experience and (2) reduction of the gains of
the visual feedback loops that control axial eye growth by pharmacological
agents. To use these approaches successfully, it has first to be clarified (1)
what the critical features in the retinal image are and (2) which messengers are
released from the retina to control the growth of the underlying sclera and how
they can be modified. Our progress in both areas will be described. In addition,
our new refraction technology will be demonstrated (the PowerRefractor). Myopia.
March 29, Noon, 489 Minor Hall
Christine Wildsoet
New England College of Optometry
Host: Stan Klein
Emmetropization, accommodation and myopia: what can the chick model tell us
about their association?
Abstract
That extensive near work and myopia are linked is a belief held by many in
relation to human myopia yet little understood in terms of causality. However,
recent research using animal models has provided some insights into the issues
involved. Progress on this front reflects, at least in part, the fact that
stimulus conditions can be more readily manipulated and controlled. In this
paper, I will present some of my more recent research using the chick as such a
model.
March 12, 12:00 Noon, 489 Minor Hall
Benjamine Backus
Stanford University
Host: Marty Banks
Response of human visual brain areas to binocular disparity
March 5, 12:00 Noon, 489 Minor Hall
Joel Miller
Smith Kettelwell Eye Research Foundation
Host: Jim Maxwell
Convergence and Co-contraction
Jan 22, 12:00 noon, 489 Minor Hall
Michael Kubovy
Psychology Dept. at Universitiy of Virginia
Fourth Annual Irvin Rock Memorial Lecture, co-sponsored with the Institute for Cognitive Studies
Host: Steve Palmer
On Perceptual Grouping
Abstract
In this lecture I will present progress we have made on understanding two
related phenomena first described close to a century ago: grouping by
proximity and grouping by similarity. I will:
-
Show how the stimuli used by Gestalt psychologists - regular dot patterns
- can be generalized and how their phenomenological approach can be extended
to give stable, replicable, and quantifiable data.
-
Present a Pure Distance Law that captures the probabilistic features of
grouping by proximity and a Attenuator Hypothesis that captures properties of
grouping by similarity.
-
Show how our data suggest a Theory of Grouping, a
probabilistic theory in which the central construct is Attraction, which
can be measured on a ratio (or even an absolute) scale.
- Sketch some open problems.
|
