Current Semester Speakers
Understanding the Remarkable Accuracy of Colour Perception
Lorne Whitehead, University of British Columbia (Physics and Astronomy)
April 20 • 11:00 am • Reynolds School of Journalism, 101
This presentation is intended primarily for non-experts in color vision and those who teach this subject. Most students are aware that colour vision arises from spectrally selective phototransduction in retinal cells. However, that picture offers little insight into how and why people usefully perceive the colours of surfaces, which was the primary evolutionary driver for color vision. With an emphasis on simple principles of physics and information processing, I will describe a recent collaborative effort that helps to explain the remarkable accuracy of the perception of surface colors. This explanation avoids complex mathematics and quantitatively models plausible physiological processes. I hope that, after this presentation, audience members will be able to describe, in general terms, the complex nature of the colour vision system, and how this complexity yields simple information that usefully characterizes the materials around us.
Studies of Sex, Eyes, and Vision: Importance of Estrogen, SWS Cones, and Lite Beer
Alvin Eisner, Portland State University (Institute on Aging/OHSU-PSU School of Public Health)
April 6 • 3:00 pm • Reynolds School of Journalism, 101
Historically, little attention has been paid to effects of hormonal change on visual function. This absence stems from many factors. Practical difficulties conducting interdisciplinary research may contribute to recalcitrant "basic" vs. "clinical" dichotomies, and tacit assumptions can lead to overgeneralization and consequent under-recognition of meaningful between-person differences. I will present data - from studies employing distinctively different subject populations - collectively showing that changes in estrogenic activity can impact vision mediated via Short-Wavelength-Sensitive (SWS) cones. SWS cones signal to the visual cortex mainly via a restricted set of neural pathways, and it has long been known that certain test/background stimulus conditions allow threshold-level incremental test stimuli to be detected via the differential response of SWS cones. Thus, it was surprising initially to find that ~1/3 of healthy postmenopausal women report that a short-wavelength test stimulus appears white at threshold rather than bluish and/or reddish, as typically experienced by men of all ages. Studies conducted subsequently suggested that the white color appearance involves sluggishness of visual response(s). SWS-cone-mediated response evidently can be affected by the selective-estrogen-receptor-modulator (SERM) tamoxifen, by aromatase inhibition (which abolishes estrogen synthesis), and even by phytoestrogen consumption. In addition, light adaptation within SWS cone pathways varies cyclically along with menstrual phase for some young women (with PMS?), and this variation can be altered by oral contraceptive use. Implications of these and other results will be discussed, e.g., as concerns breast cancer survivorship.
Duck! How Your Brain Works Out 3D Motion from 2D (Retinal) Sensory Signals
Lawrence K. Cormack, University of Texas at Austin (Psycholoy/Neuroscience)
March 30 • 11:00 am • Reynolds School of Journalism, 101
Animals, by definition, move. For animals to interact with one another and to cope with their own animation, they must be able to correctly perceive movement, especially relative movement towards the animal itself (3D motion). For creatures of light and air, such as ourselves, this is accomplished through a binocular visual system. Visual systems, however, employ eyeballs, which have an inherently 2D sensor surface. The brain thus faces the challenge of recovering estimates of motion through a 3D environment given an input that lacks any explicit coding of the 3rd dimension. Over the past decade or so, using a coordinated combination of behavioral experiments, functional magnetic resonance imaging, and cortical electrophysiology, we have learned a great deal about how the brain meets this challenge. The end result compels a surprising and substantial reformulation of the "standard model" of cortical motion processing in primates.
How the Olfactory Experience Sculpts the Olfactory System
Stephen Santoro, University of Wyoming (Zoology & Physiology)
March 14 • 4:00 pm • Davidson Math & Science
Computational and Cortical Modeling of Lightness and Color Perception
Michael E. Rudd, University of Washington (Physiology & Biophysics)
February 23 • 11:00 am • Reynolds School of Journalism, 101
I will describe some psychophysical experiments conducted in my lab to study effects of spatial context on the lightness and color in simple visual displays. The perceptual results will then be combined with neurophysiological data from other labs to motivate a computational theory of cortical color processing. The theory assumes that spatially-directed luminance change at luminance edges, and within luminance gradients, are computed in cortical areas V1 and V2, then the relevant neuronal outputs are spatially integrated at a higher stage of cortical processing (probably in area V4, TEO, or TE) to compute surface color. The model architecture suggests that the neural correlate of surface color percepts arises late in the cortical ventral pathway, near the areas associated with object perception. The neural computations assumed by the model share algorithmic properties with Land's early Retinex color vision model, which was designed to achieve color constancy under the challenge of changes in overall illumination. However, the cortical model is considerably more complex than Retinex in that it incorporates additional properties of visual neural processing, including top-down influences (attention), neural analysis at different spatial scales, and different neural gains of ON- and OFF- cells. To help illustrate the behavior of the model, I will demonstrate how it explains various visual illusions-including classical contrast and assimilation phenomena, brightness and color filling-in, and a new illusion in which surrounding a light or dark patch with a luminance gradient can reverse the perceived contrast polarity of the patch-as results of the misapplication to artificial stimuli of biological computations that evolved to support color constancy under natural viewing conditions.
Reasoning with Uncertainty the Bayesian way with examples in Cognitive Modeling in R and Stan
A. Grant Schissler, University of Nevada, Reno (Mathematics & Statistics)
February 9 • 11:00 am • Reynolds School of Journalism, 101
Scientific discovery and learning from data are challenging tasks. Understanding uncertainty through statistical modeling is essential in all areas of research. However, widely-used statistical constructions including the infamous P-value often do more harm than good in the pursuit of knowledge. Indeed, P-values are under attack in both statistical and domain-specific communities. In this talk, I'll discuss some problems in P-value decision theoretic reasoning and present a flexible and philosophical coherent strategy - Bayesian modeling and inference. Recent advances in computation and software provide extremely fast and relatively straightforward implementation of complex Bayesian models. After providing background on Bayesian modeling in general, we'll walk through some examples in cognitive modeling, such as inferring IQ scores using Gaussian processes, hierarchical signal detection, and psychophysical functions. The case studies will be demonstrated in R and Stan, and code will be provided to serve as templates. The session will conclude with ample time for discussion.
Top-down and Bottom-up Modulation of Neural Coding in the Somatosentory Thalamus
Qi Wang, Columbia University (Biomedical Engineering)
December 15 • 11:00 am • Reynolds School of Journalism, 101
The transformation of sensory signals into spatiotemporal patterns of neural activity in the brain is critical in forming our perception of the external world. Physical signals, such as light, sound, and force, are transduced to neural electrical impulses, or spikes, at the periphery, and these spikes are subsequently transmitted to the neocortex through the thalamic stage of the sensory pathways, ultimately forming the cortical representation of the sensory world. The bottom-up (by external stimulus properties) or top-down (by internal brain state) modulation of coding properties of thalamic relay neurons provides a powerful means by which to control and shape information flow to cortex. My talk will focus on two topics. First, I will show that sensory adaptation strongly shapes thalamic synchrony and dictates the window of integration of the recipient cortical targets, and therefore switches the nature of what information about the outside world is being conveyed to cortex. Second, I will discuss how the locus coeruleus - norepinephrine (LC-NE) system modulates thalamic sensory processing. Our data demonstrated that LC activation increased the feature sensitivity, and thus information transmission while decreasing their firing rate for thalamic relay neurons. Moreover, this enhanced thalamic sensory processing resulted from modulation of the dynamics of the thalamorecticulo-thalamic circuit by LC activation. Taken together, an understanding of the top-down and bottom-up modulation of thalamic sensory processing will not only provide insight about neurological disorders involving aberrant thalamic sensory processing, but also enable the development of neural interface technologies for enhancing sensory perception and learning.
Adaptation to the Variability of Visual Information
John Maule, University of Sussez (Psychology)
December 5 · 12:00 pm • Schulich Lecture Hall 3
The sensory signals we can detect from the world are highly variable and the brain has a large amount of information to process, encode and represent percepts. One way in which the visual system can reduce its processing load is to use summary statistics - representing the mean of features in the set, rather than individual exemplars. It has previously been found that observers are able to extract the mean hue from a rapidly-presented ensemble of colours (e.g. Maule & Franklin, 2016). This ability has been demonstrated for other stimulus domains, including orientation, size and facial expression. In addition to summary statistics of central tendency, it may also be useful for the visual system to encode information about the variation present in visual features. I will present a series of experiments investigating the encoding of variance for colourful ensembles. Ensemble variance was controlled by varying the difference in hue (in CIELUV colour space) between different elements. Observers viewed pairs of ensembles situated to the left and right of a central fixation point. During the adaptation phase there was a consistent relationship between the amount of variance in each ensemble (e.g., left more variable in hue than right). On test trials observers judged which ensemble appeared more variable. Generally, following exposure to highly variable ensembles on the left observers perceived a pair of equally variable ensembles as relatively less variable on the left compared to the right of the display. This result is similar to that shown by Norman et al. (2015) for ensembles of orientation, suggesting that representation of the variance independent of the central tendency may be a general feature of visual coding. The results imply that perceived variability of a multi-coloured ensemble is subject to adaptation after-effects, and therefore that colour variance is an encoded property of visual sets. The value of encoding variability may be in tuning the brain to the visual properties of the immediate surroundings, allowing the brain to better predict the content of the environment and represent salient elements.
Spatial Vision at the Scale of the Cone Photoreceptor Mosaic
David Brainard, University of Pennsylvania (Psychology)
November 17 • 12:00 pm • Reynolds School of Journalism 101
The long-term goal of the research here is to understand how the visual system integrates information from individual cones in the photoreceptor mosaic, to produce the high-resolution percept of a colored world that we enjoy. A particular richness of this question derives from the observation that there are three distinct spectral classes of cones in the mosaic, and that cones of these classes are arranged in an interleaved fashion. To understand how signals from individual cones are combined, we have begun to employ adaptive optics retinal imaging together with real time eye tracking to conduct psychophysical experiments. In these experiments, stimuli whose scale approaches that of individual cones are targeted to precisely defined retinal locations, and we measure detection thresholds as the spatial structure of the stimuli is varied. In this talk, I will describe our methods along with initial results that examine spatial summation for human foveal vision, with the smallest stimuli matched in size to the acceptance aperture of a single cone.
Action Selection According to Ideomotor Theory: Basic Principles and an Application to Multitasking
Markus Janczyk, Eberhard Karls University of Tübingen (Psychology)
November 6 · 11:30 am • Reynolds School of Journalism 101
Human act to achieve certain goals. Ideomotor Theory, advanced in the 19th century by several philosophers, claims that actions can only be addressed by mentally anticipating the desired goal states. This idea was rarely investigated in psychology for a long time, but during the last decades several lines of empirical investigation were pursued with results being in line with Ideomotor Theory.
I will begin this talk with an introduction into Ideomotor Theory and the main evidence from recent studies, followed by a brief introduction into dual-tasking models. I will then bring these fields together and sketch several lines of research investigating the role of goal anticipations for dual-task performance. In sum, the (1) capacity-limited stage of processing - assumed to be the cause of dual-task problems - can be described as goal anticipation, the (2) commensurability of goal states affects the amount of dual-task problems, and (3) monitoring the occurrence of pursued goal states also incurs costs. These results suggest that anticipating goal states is an important contribution to dual-task problems.
Multidimensional Estimation of Color Matching Functions
November 3 · 1:00 pm • Reynolds School of Journalism 101
For many industrial and academic applications, it is essential to know the color responses of observers to arbitrary scene spectral radiances. The spectral response of an observer can be defined as the detected quantum efficiency resulting from radiation of a given wavelength, over the range of all wavelengths to which the observer is sensitive. These spectral responses are commonly referred to as spectral sensitivities or more generally as color matching functions. In the case where the observers are modern digital cameras, an important step in the color image processing pipeline is the transformation from camera response to objective colorimetric or related quantities, often for each individual unit. In the case where the observers are persons, human color matching functions also map scene spectra to colorimetric quantities, though every person has somewhat different color matching functions. One application of widespread interest is soft proofing where various personnel must ensure colorimetric or appearance matches between images viewed on various wide-gamut displays and prints viewed in controlled light-booths. Since modern displays are often based on narrow-band primaries, variations in individual human color matching functions and display primary spectra can cause significant color matching errors. For many critical color matching applications, there is widespread interest in how best to determine individual color matching functions. However, direct determination of individual color matching functions in these settings is tedious and impractical for many reasons. In this presentation, a method and results are shown that allow accurate estimation of camera spectral sensitivities based on a few simple measurements and the case is made for extending the method to estimating color matching functions for individual human observers.
Symposium - 9th Annual, Sierra Nevada Chapter of the Society for Neuroscience (pdf)
October 26 • 12:00 pm-5:00pm • Pennington Health and Science 102
Clark Elliott, DePaul University (Institute of Applied Artificial Intelligence)
September 22 • 3:00 pm • Joe Crowley Student Union Ballroom A
The brain is primarily a visual-spatial processing device. This has implications for all aspects of human cognition and sensory interpretation. Neurodevelopmental optometry accesses the brain through the retinas, giving us a high-bandwidth mechanism for measuring many of the kinds cognitive brain function that are at the core of what makes us human, as well as for altering such brain configurations by taking advantage of the brain's plastic nature. In this talk I will present a self-reporting case study of a ten-year odyssey with significant brain dysfunction resulting from an mTBI - including "permanent" impairments such as balance difficulties, inability to initiate action, inability to read, to walk, to understand speech, and to sleep normally-but ultimately ending with the truly rare full recovery after brain reconfiguration using neurodevelopmental optometric techniques. In the second part of the talk we will look at the three retinal pathways that neurodevelopmental optometry treats, including center vision, peripheral processing, and a collection of critical non-visual retinal pathways that have been all but ignored until recently, and which are often of great importance in treating brain injuries. We will also discuss a set of clinical qEEG scans showing the changes in brain activity when wearing therapeutic eyeglasses.
The Adult Face-Diet Revealed: Impact of Daily Face Exposure on the Perception of Faces
Ipek Oruc, University of British Columbia (Ophthalmology & Visual Sciences)
July 21 • 11:00 am • Reynolds School of Journalism 101
Faces are ecologically significant stimuli central to social interaction and communication. Human observers are considered to be experts in face perception due to their remarkable ability to recall great numbers of unique facial identities encountered in a lifetime, their sensitivity to subtle differences that distinguish different identities, and their robustness across significant differences among images of the same identity. A large body of work in the last several decades have investigated limits to this expertise such as recognition of faces of unfamiliar races ("the other-race effect") and faces viewed in the inverted orientation ("the face-inversion effect"). In this talk, I will describe recent results from our group that have suggested that face size, as a proxy for viewing distance, impacts face recognition processing and performance. Furthermore, I will present results from our recent naturalistic observation study that examined adults' daily face exposure, i.e., the adult face-diet. I will compare the adult face-diet to what is known about that of infants and consider these results in light of effects of size on face recognition. I will speculate about origins of these size effects and consider contributions from innate and genetic factors, early exposure during sensitive periods of development, and late exposure during adulthood.
Selectivity, Hyper-selectivity, and a General Model of the Non-linearities of Neurons in the Visual Pathway
David Field, Cornell University (Psychology)
July 18 • 12:00 pm • Reynolds School of Journalism 101
I will discuss some implications of an approach that attempts to describe the various non-linearities of neurons in the visual pathway using a geometric framework. This approach will be used to make a distinction between selectivity and hyper-selectivity. Selectivity will be defined in terms of the optimal stimulus of a neuron, while hyper-selectivity will be defined in terms of the falloff in response as one moves away from the optimal stimulus. With this distinction, I show that it is possible for a neuron to be very narrowly tuned (hyper-selective) to a broadband stimulus. We show that hyper-selectivity allows V1 neurons to break the Gabor-Heisenberg localization limit. The general approach will be used to contrast different theories of non-linear processing including sparse coding, gain control, and linear non-linear (LNL) models. Finally, I will show that the approach provides insights into the non-linearities found with overcomplete sparse codes - and argues that sparse coding provides the most parsimonious account of the common non-linearities found in the early visual system.
Development of Neural Mechanisms Underlying Face Recognition Ability
Vaidehi S. Natu, Stanford University (Psychology)
July 7 • 11:00 am • Reynolds School of Journalism 101
Human face recognition ability is critical for social interactions and communication and it improves from childhood to adulthood. Face-selective regions in the ventral stream increase in size and neural responses to faces, and these developments are related to better face recognition ability. However, it is unknown whether these developments affect perceptual discriminability of faces and whether they are accompanied with anatomical changes. My talk will describe my research addressing these open questions. First, I will describe results of an fMRI-adaptation study, conducted in children (ages 5-12) and adults (ages 22-28), that was aimed to determine if neural sensitivity to faces develops. Our data shows that neural sensitivity to face identity in face, but not object-selective cortex, develops with age, and this development is correlated with increased perceptual discriminability for faces. Second, I will present results from a study that investigated neural mechanisms of anatomical development in face-selective regions. While gray matter across ventral temporal cortex thins from age 5 to adulthood, it is unknown if this thinning is due to pruning or increased myelination. Using novel quantitative MRI and diffusion MRI techniques I will present new evidence that tissue growth and myelination in deep cortical layers, not pruning, is associated with cortical thinning. Together these new data elucidate the functional and anatomical development of face-selective regions from childhood to adulthood and provide an important foundation to understand typical and atypical development.