25th Symposium of the University of Rochester Rochester ... › media › pdf › symposium ›...

25th Symposium of the Center for Visual Science University of Rochester Rochester, NY 14627

Statistical Learning and

Brain Plasticity

May 31 - June 3, 2006

Organizing Committee Richard Aslin - Daphne Bavelier - Alexandre Pouget

Symposium Secretary

Debbie Shannon University of Rochester (585) 275-2459 Voice (585) 271-3043 Fax

Symposium sponsored by the Office for Naval Research

GENERAL INFORMATION Free round-trip shuttle transportation will be provided from the hotels to the

Rochester International Airport and to the University of Rochester River Campus (Wednesday) and Medical Center (Thursday-Saturday).

Meal tickets for the lunches and dinners are in your registration packet if requested on

your registration form. A separate lunch line will be provided for vegetarians. Please take one serving per person when going through the food line to ensure that everyone receives his or her meal. After everyone has been served, the remainder of the food will be available for general consumption.

On Saturday, an announcement will be made during the final session to orchestrate

transportation to the banquet. Carpooling will be organized outside the registration area at the medical center. Parking will be available at the George Eastman House. The dinner will be held directly after the final sessions, so dress is casual. There will be open bar during the Saturday reception and wine will be served with dinner.

There will be a public wireless network in the Medical Center Atrium called "CVS".

No username/password should be necessary to connect. If you have difficulties, contact the registration desk.

Please remember to turn off all cell phones, PDAs, and pagers during the talks.

Organizing Committee: Richard Aslin Daphne Bavelier Alexandre Pouget

Sponsored by the Office for Naval Research

POSTER PRESENTATION INSTRUCTIONS The poster should be put up early Friday morning, June 2. The poster session will be

held in the Flaum Atrium, which is adjacent to the meeting room. You are asked to stay with your poster from 2:30 - 4:30 pm on Friday afternoon. You should take down your poster after the session has concluded, at 4:30 pm.

The posters should be prepared to fit on an approximately 5 ft x 3.5 ft foam core

poster board that has a white background. All boards will be numbered. Attach your poster to the board that corresponds to

your number on the "Poster Presentations" list in the program booklet. Bring pushpins for attaching your poster. Do not use tape. We will have a limited

supply of pushpins available at the registration desk should you be unable to bring your own.

Do not try to mount heavy materials, as they will have difficulty staying attached to

the foam core poster board. Please see the administrative staff at the registration desk if you have any difficulties.

Program Schedule May 31 - June 3, 2006

*All talks and discussion sessions will be held in the Class of '62 Auditorium, University of Rochester Medical

Center

Wednesday, May 31

4:00 pm - 7:00 pm Registration and Welcome Reception, 269 Meliora Hall

Thursday, June 1

8:30 am Registration & Breakfast, Medical Center Atrium

9:00 am Welcome, David Williams

Perceptual and Motor Learning, David Williams, Session Chair 9:10 am - 9:50 am Mario Svirsky Indiana University

Learning to understand frequency-shifted, spectrally

degraded speech (T1)

9:50 am - 10:30 am Jason Gold Indiana University

Signal and noise in perceptual learning (T2)

10:30 am - 11:00 am Break (Refreshments will be served), Medical Center Atrium

11:00 am - 11:40 am Reza Shadmehr Johns Hopkins University

Internal models, adaptation, and uncertainty (T3)

11:40 am - 12:20 pm Alexandre Pouget University of Rochester

Neural basis of perceptual learning: haven't we solved

this issue already? (T4)

12:20 pm - 12:40 pm Discussion Session

12:40 pm - 2:00 pm Lunch, Medical Center Atrium

Learning: Role of Priors and Attention, Robbie Jacobs, Session Chair

2:00 pm - 2:40 pm David Knill University of Rochester

Learning Bayesian priors for depth perception (T5)

2:40 pm - 3:20 pm Marvin Chun Yale University

Attentional control of perceptual memory (T6)

3:20 pm - 3:50 pm Break (Refreshments will be served), Medical Center Atrium

3:50 pm - 4:30 pm Nick Chater University College London

Simplicity, probability and perception (T7)

4:30 pm - 5:10 pm Josh Tenenbaum Massachusetts Institute of Technology

Statistical learning of abstract knowledge (T8)


6:00 pm - 7:30 pm Dinner, Medical Center Atrium

Friday, June 2 9:00 am Breakfast, Medical Center Atrium

Constraints on Pattern Learning, Michael Weliky, Session Chair

9:30 am - 10:10 am Lori Holt Carnegie Mellon University

Auditory categorization and tuning in speech perception

(T9)

10:10 am - 10:50 am Daniel Margoliash University of Chicago

Pattern perception in songbirds (T10)


11:20 am - 12:00 pm Richard Aslin University of Rochester

Statistical learning of visual patterns: Helmholtz, Bayes,

and Dr. Spock (T11)

12:00 pm - 12:40 pm Toby Mintz University of Southern California

Learning syntactic categories from patterns in linguistic

input (T12)



2:00 pm Tour of the Center for Brain Imaging, Medical Center Annex (optional)

Afternoon Poster Session

2:30 pm - 4:30 pm Poster Session, Informal Discussions, Medical Center Atrium

Free Evening

Saturday, June 3

8:30 am Breakfast, Medical Center Atrium

Neural Mechanisms of Learning, Lizabeth Romanski, Session Chair

9:00 am - 9:40 am Takao Hensch Riken Institute

Title TBA (T13)

9:40 am - 10:20 am Anthony Zador Cold Spring Harbor Laboratory

How many synapses must change to form a memory?

(T14)


10:50 am - 11:30 am Nathaniel Daw University College London, Gatsby

Reward & exploration in human decision making (T15)

11:30 am - 12:10 pm Leo Sugrue Stanford University

Choosing the greater of two goods: a combined

behavioral, modeling, and physiological approach to

value based decision making (T16)



Maturation and Plasticity, Daphne Bavelier, Session Chair

2:00 pm - 2:40 pm Daphne Maurer McMaster University

Missed sights: consequences for visual development

(T17)

2:40 pm - 3:20 pm Brian Wandell Stanford University

Maps and reading development in visual cortex (T18)

3:20 pm - 3:50 pm Break (Refreshments will be served), Medical Center Atrium

3:50 pm - 4:30 pm Elissa Newport University of Rochester

Statistical language learning: computational and

maturational constraints (T19)


4:50 pm - 5:10 pm Break

5:10 pm - 6:00 pm Randy Gallistel Rutgers University

Is mutual information the learning-relevant parameter of

conditioning protocols? (T20)

6:00 pm - 6:30 pm General Discussion

7:00 pm - 9:30 pm Dinner, George Eastman House

End of Meeting

POSTER PRESENTATIONS (P1) BACKUS, B: QUANTIFICATION OF THE VISUAL SYSTEM'S RELIANCE ON A NEWLY RECRUITED CUE USING CUE CONFLICT STIMULI (P2) BECK, J: BAYESIAN INFERENCE WITH PROBABILISTIC POPULATION CODES (P3) BROUWER, A: USING MEMORIZED AND VISUAL LOCATION INFORMATION IN GOAL DIRECTED MOVEMENTS (P4) CHUKOSKIE, L: THE INFLUENCE OF PRIOR EXPERIENCE ON SACCADE CHOICE (P5) FRANK, M: INTEGRATING VISUAL INFORMATION INTO PROBABLISTIC MODELS OF WORD LEARNING (P6) GREEN, C: ABILITY TO TASK-SWITCH IN ACTION VIDEO GAME PLAYERS (P7) KELLY, K: TRAINING-INDUCED IMPROVEMENTS IN MOTION SENSITIVITY AFTER V1 DAMAGE IN HUMANS (P8) MCKEEFF, T: ATTENTION CAN ALTER THE TEMPORAL CAPACITY OF OBJECT PROCESSING IN HIGH-LEVEL VISUAL AREAS (P9) PADILLA, M: DOES PERCEPTION LEARNING ACQUIRE THE PRIOR? (P10) RILEY, M: HOMONYMOUS HEMIANOPIA ALTERS THE DISTRIBUTION OF FIXATIONS IN 3D VIRTUAL ENVIRONMENT (P11) SOHN, J: ACTION VALUE IN THE SUPPLEMENTARY AND PRESUPPLEMENTARY MOTOR AREAS (P12) WONNACOTT, E: ACQUIRING AND PROCESSING VERB ARGUMENT STRUCTURES: A MINIATURE LANGUAGE STUDY (P13) YOON, J: THE DEVELOPMENT OF OBJECT AND FACE PROCESSING IN CHILDREN

25th Symposium: Speaker Abstracts

Center for Visual Science

Perceptual and Motor Learning

T1 May 31 - June 3, 2006

LEARNING TO UNDERSTAND FREQUENCY-SHIFTED, SPECTRALLY DEGRADED SPEECH Mario Svirsky, Indiana University The human brain is a remarkable speech recognizer, even in the face of such extreme distortions as sinewave speech, clipped speech, and four-channel noise vocoders. However, the simultaneous application of two types of distortion, such as a 1 or 2 octave frequency shift combined with the spectral degradation inherent in an 8-channel noise vocoder, renders the input speech signal unintelligible. This has important clinical implications because this type of combined distortion is a reasonable model of the speech input received by cochlear implant (CI) users. Fortunately, it is possible to learn how to interpret such a signal so that it becomes more intelligible over time. It is generally believed that CI’s impose a basalward shift to the acoustic input, that is, sounds stimulate neurons with higher characteristic frequency than the acoustic frequency of the original stimulus. This frequency misalignment may have a negative influence on speech perception by postlingually deaf CI users. However, a perfect frequency-place alignment between analysis filters and stimulated electrodes may result in the loss of important low frequency speech information. A trade-off may involve a gradual approach: start with correct frequency-place alignment to allow listeners to adapt to the spectrally degraded signal first, and then gradually increase the basalward shift to allow them to adapt to it over time. A first study compared the gradual approach described above to the sudden approach normally used with CI users: immediate exposure to a constant frequency shift that does not change over time. The main finding was that speech perception scores were initially much higher with the gradual approach than with the sudden approach, but differences decreased over the course of 15 one-hour training sessions. A second study also employed spectrally degraded, frequency shifted speech, but listeners were allowed to adjust the input filter bank in real time to a preferred setting. Interestingly, the listener-selected filter banks represented a tradeoff between correct frequency-place alignment and low frequency speech information. Taken together, these results may have significant implications for the optimal fitting of sensory aids such as cochlear implants and frequency transposition hearing aids.




T2 May 31 - June 3, 2006

SIGNAL AND NOISE IN PERCEPTUAL LEARNING Jason Gold, Indiana University Performance in perceptual tasks often improves dramatically with practice ('perceptual learning'). In this talk, I will discuss recent work directed at specifying the mechanisms that mediate perceptual learning, cast within the framework of signal detection theory and black-box perceptual information processing models. Within this context, I will also discuss a collection of system-identification techniques that use externally added noise to quantify the factors that limit performance and allow one to trace how these factors change as a function of practice.




T3 May 31 - June 3, 2006

INTERNAL MODELS, ADAPTATION, AND UNCERTAINTY Reza Shadmehr, Johns Hopkins University When the brain generates a motor command, it also predicts the sensory consequences of the motor commands in terms of how it will affect the environment, and the body. What is the purpose of this prediction? Here I will show that the brain integrates its predictions with the actual sensory feedback, arriving at an estimate that is better than possible from sensation alone. This Bayesian integration depends on sensorimotor maps that must are models of our body and the environment. Because these systems can change, the maps must adapt. I will show that a prediction error causes changes in multiple adaptive systems. Some are highly responsive to error, but rapidly forget. Others are poorly responsive to error but have high retention. This explains savings and spontaneous recovery. Fast and slow adaptive processes arose because disturbances to the motor system have various timescales (fatigue vs. disease). When faced with error, the brain faces a credit assignment problem: what is the timescale of the disturbance? To solve this problem, the brain likely keeps a measure of uncertainty about the timescales.




T4 May 31 - June 3, 2006

NEURAL BASIS OF PERCEPTUAL LEARNING: HAVEN'T WE SOLVED THIS ISSUE ALREADY? Alexandre Pouget, University of Rochester Extensive training on simple perceptual tasks often improves behavioral performance. The neural basis of this improvement is believed to be well-understood. It is thought to involve a combination of the following changes: 1- an increase in the number of neurons representing the sensory input, 2- a sharpening of the tuning properties of the sensory neurons, and 3- an increase in the average firing rates. Modeling studies have suggested that all three of those changes could indeed increase the information content of cortical population codes, which in turn could account for the improved performance. These studies however are all based on the assumption that the noise in the brain is independent. Unfortunately, this assumption is incorrect: neurons are correlated and, importantly, correlations are likely to change during learning. I will show that when correlations are taken into account, a mechanism such as tuning curve sharpening can either increase or decrease information depending on how it is implemented. Moreover, it is possible to increase the information content of a population code by adjusting correlations while leaving intact the tuning curves. Therefore, the neural basis of perceptual learning is still very much unclear, but could be resolved by using multielectrode recordings.



Learning: Role of Priors and Attention

T5 May 31 - June 3, 2006

LEARNING BAYESIAN PRIORS FOR DEPTH PERCEPTION David Knill, University of Rochester Pictorial cues to depth rely on prior knowledge about statistical regularities in the environment; for example about the prevalence of symmetric objects, of parallel lines and lighting from above. In the first part of this talk, I will discuss evidence that the human visual system uses probabilistic characterizations for this prior knowledge that incorporate mixtures of multiple possible models of objects. This account explains a number of perceptual effects, most notably non-linear robust cue integration. In the second part of the talk, I will discuss the problem of how the visual system learns the statistical regularities needed to interpret pictorial cues to; in particular, how it adapts its internal model to environments with very different statistics. We specifically tested the hypothesis that the visual system can adapt its model of the statistics of planar figures for estimating 3D surface orientation. Taking elliptical figures as a prototypical case, we develop a Bayesian model that effectively learns the probability density function on shape from stereoscopic images of slanted ellipses. When the model adapts to an irregular environment, it gradually down-weights the pictorial cue to slant provided by the shapes of projected ellipses relative to stereopsis. When estimating surface slant in an environment containing randomly shaped ellipses, human subjects similarly down-weight the pictorial cue over time, but not in an environment containing mostly circles. This shows that they have adapted their internal model of the shape statistics of the environment.




T6 May 31 - June 3, 2006

ATTENTIONAL CONTROL OF PERCEPTUAL MEMORY Marvin Chun, Yale University Functional magnetic resonance imaging (fMRI) reveals not only where information is processed in the brain, but also what is encoded in a stimulus-specific manner. For example, just as infants, children, and adults habituate to events that they treat as "the same," underlying neural responses measured with fMRI are lower to repeated, familiar stimuli than to novel stimuli. This difference is known as repetition attenuation, and such attenuation signals allow researchers to study stimulus-specific perceptual and memory representations in the brain (Grill-Spector et al., 2006; Schacter & Buckner, 1998). Prior work has focused on linking this repetition attenuation signal with automatic perceptual processing and unconscious, implicit memory (Wiggs & Martin, 1998). In contrast, our work indicates that repetition attenuation is subject to attentional control and also correlated with conscious, explicit memory (Turk-Browne et al., 2006; Yi & Chun, 2005). Having established a strong association between repetition attenuation and perceptual memory, a final study will use repetition attenuation to reveal systematic distortions, that is, false memories, of scene layout information that was never seen by the observer (Park et al., submitted).




T7 May 31 - June 3, 2006

SIMPLICITY, PROBABILITY AND PERCEPTION Nick Chater, University College London Perception is a problem of abductive inference—inferring the structure of the environment, from patterns of stimulation on the sensory surfaces. Abductive inference can naturally be modelled in as Bayesian framework—but this requires assigning prior probabilities to what appears to an infinite range of perceptual hypotheses. How is this possible? This talk suggests that priors can usefully be viewed as set by the choice of representation 'language' in which perceptual input is coded. This leads to a simplicity principle in perception—that is, the perceptual system is viewed as preferring short codes for sensory input. This approach is part of a long tradition in the study of perception, dating back to Mach and the Gestalt psychologists. This 'simplicity' perspective on perception provides a natural interpretation of the Gestalt Laws, and a range of phenomena of perceptual organization. I will also consider the question of the scope and limitations of this approach, and discuss the question of how far this framework is open to empirical test.




T8 May 31 - June 3, 2006

STATISTICAL LEARNING OF ABSTRACT KNOWLEDGE Josh Tenenbaum, Massachusetts Institute of Technology In accounts of cognitive development, statistical learning is generally seen as an alternative to the acquisition and use of abstract knowledge and richly structured representations. Empiricist typically embrace the former while shying away from the latter; rationalists or nativists typically take the opposite stance. Yet these two concepts—statistical inference and abstract, structured representation—are two of the most powerful tools that have ever been offered to explain the nature and origins of intelligence, in over two thousand years of trying. Must we be forced to choose between them? I will argue that theories of learning and development can and must draw on the strengths of both structure and statistics. I will present a hierarchical Bayesian framework for inductive learning, in which statistical inference operates over structured representations of knowledge at multiple levels of abstraction. These knowledge representations may be thought of as simple forms of intuitive theories for various domains of entities, properties, and relations. The hierarchical Bayesian analysis shows how abstract knowledge of domain structure provides strong constraints on a learner's inductive generalizations, and how that abstract knowledge may itself be learned through rational statistical means. I will discuss applications of the framework to modeling learning and reasoning in several domains, such as natural kind categories and their properties, social relations or causal relations. (Joint work with Charles Kemp, Vikash Mansinghka, and Tom Griffiths.)



Constraints on Pattern Learning

T9 May 31 - June 3, 2006

AUDITORY CATEGORIZATION AND TUNING IN SPEECH PERCEPTION Lori Holt, Carnegie Mellon University The ease of everyday conversation masks the cognitive and perceptual challenges of translating from acoustic signal to meaning. One of the fundamental reasons machine recognition of speech is so difficult is that the acoustics of spoken language are incredibly complex. The acoustic details of a spoken utterance vary with the rate of speech, speaker idiosyncrasies, phonetic context, accent, and even the reverberance of the speaking environment. With these diverse sources of acoustic variability—some linguistically relevant, some not—the mechanisms that transform the acoustic signal into a linguistic representation face a complex task. However, the acoustic speech signal and the greater perceptual environment in which it is presented also possess much regularity. I will present data demonstrating that some of the perceptual challenges of speech perception, and thereby some of the challenges of early language processing, may be met by general-purpose perceptual mechanisms sensitive to regularity and change in the perceptual environment at multiple time scales.




T10 May 31 - June 3, 2006

PATTERN PERCEPTION IN SONGBIRDS Daniel Margoliash, University of Chicago Temporal sequence is a rich and essential component of information in vocalizations. As part of a larger effort to characterize nonlinear receptive field properties of higher-order auditory neurons we have been studying sensitivity of starlings to sequences of naturally-occurring motifs. After achieving baseline performance on a go/nogo task that contrasted sets of strings drawn from finite-state and context-free grammars (CFG), birds were transferred to novel sets of strings of the same order, and probed with higher-order grammatical strings and agrammatical strings. The results indicate that birds learned a simple CFG, a level of perceptual syntactic complexity which has previously been posited to be uniquely available to humans. Starlings performed equally well on strings of human syllables whereas humans easily solved the problem when posed with strings of syllables but struggled with strings of motifs. These results emphasize the importance of species specificity in all animal behavior including language and challenge dogma that places language outside the realm of biological experimentation. Higher-order neurons in the starling auditory system respond selectively to motifs depending on learning and reward contingencies. Although we know little yet regarding temporal sequence sensitivity of these neurons, such analysis is likely to provide mechanistic insight into complex pattern perception in birds, which may constrain theories of the evolution of such behaviors.




T11 May 31 - June 3, 2006

STATISTICAL LEARNING OF VISUAL PATTERNS: HELMHOLTZ, BAYES, AND DR. SPOCK Richard Aslin, University of Rochester The natural visual environment is filled with objects and surfaces whose elements and their spatial arrangement form a myriad of possible underlying structures, only a subset of which correspond to the physical properties of the 3D world. How does a naive learner exposed to such a complex set of inputs manage to extract the "right" structure in finite time? A series of experiments using arrays of simple shapes organized into "scenes" will be reviewed. These experiments show that adult learners can rapidly and efficiently extract the underlying structure of these scenes without feedback (i.e., by mere exposure). Despite the presence of many spurious coincidences, a variety of constraints enable this process of statistical learning to succeed. These empirical data are well described by a class of Bayesian models (Sigmoid Belief Networks). From where do the constraints that enable statistical learning originate? Infants do not have access to top-down information and must either have intrinsic biases or acquire them from visual experience. These priors serve to make a seemingly intractable learning problem remarkably efficient, even for 9-month-olds.




T12 May 31 - June 3, 2006

LEARNING SYNTACTIC CATEGORIES FROM PATTERNS IN LINGUISTIC INPUT Toby Mintz, University of Southern California Grammatical categories—such as noun, verb, adjective, etc.—are the building blocks of syntactic structure. A crucial question in language acquisition research is how learners initially categorize words. One theory proposes that learners attend to lexical co-occurrence patterns, noting the environments in which words occur, and categorizing words together that can occur in similar environments. For example, noting that 'cat' and 'dog' both can occur after 'the' and 'a', or before 'runs', etc., would cause a learner to categorize them together. Research has shown that the linguistic input to children is structured such that distributional information of this type is indeed informative for categorizing words (Mintz et al., 2002; Redington et al., 1998). However, many behavioral studies have failed to find evidence that human learners use this kind of information on its own (e.g., Smith, 1966). I will discuss a particular sequential pattern called a frequent frame (Mintz, 2003), and show that it provides robust, cross-linguistically available cues to categories, and requires minimal computational resources. Further, I will present behavioral research with adults (Mintz, 2002) and infants (Mintz, 2006) suggesting that human learners are, in fact, especially sensitive to this kind of pattern and use it to categorize words.



Neural Mechanisms of Learning

T13 May 31 - June 3, 2006

Takao Hensch, Riken Institute




T14 May 31 - June 3, 2006

HOW MANY SYNAPSES MUST CHANGE TO FORM A MEMORY? Anthony Zador, Cold Spring Harbor Laboratory To elucidate molecular, cellular, and circuit changes that occur in the brain during learning, we investigated the role of a glutamate receptor subtype in fear conditioning. In this form of learning, animals associate two stimuli, such as a tone and a shock. Here we report that fear conditioning drives AMPAtype glutamate receptors into the synapse of a large fraction of postsynaptic neurons in the lateral amygdala, a brain structure essential for this learning process. Furthermore, memory was reduced if AMPA receptor synaptic incorporation was blocked in as few as 10% to 20% of lateral amygdala neurons. Thus, the encoding of memories in the lateral amygdala is mediated by AMPA receptor trafficking, is widely distributed, and displays little redundancy.




T15 May 31 - June 3, 2006

REWARD AND EXPLORATION IN HUMAN DECISION MAKING Nathaniel Daw, University College London, Gatsby We have rather detailed, if tentative, information about how organisms learn from experience to choose better actions. But it is much less clear how they arrange to obtain this experience. The problem of sampling unfamiliar options is a classic theoretical dilemma the costs and benefits of exploring unfamiliar options must be balanced against those of exploiting the options that appear best on current knowledge. Using behavioral analysis and functional neuroimaging, we study how humans approach this dilemma in a free-choice decision task. We assess the fit to participants' trial-by-trial choices of different exploratory strategies from reinforcement learning, and, having validated an algorithmic account of behavior, use it to infer subjective factors such as when subjects are exploring versus exploiting. These estimates are then used to search for neural signals related to these phenomena. The results support the hypothesis that exploration is encouraged by the active override of an exploitative choice system, rather than an alternative, computationally motivated hypothesis under which a single (putatively dopaminergic) choice system integrates information about both the exploitative and exploratory values of candidate actions. Although exploration is ubiquitous, it is also difficult to study in a controlled manner: We seize it only through the tight integration of computational, behavioral, and neural methods.




T16 May 31 - June 3, 2006

CHOOSING THE GREATER OF TWO GOODS: A COMBINED BEHAVIORAL, MODELING, AND PHYSIOLOGICAL APPROACH TO VALUE BASED DECISION MAKING Leo Sugrue, Stanford University To forage successfully animals must learn and maintain an internal representation of the value of competing options and link that representation to the neural processes responsible for decision-making and motor planning. To explore the neural substrate of valuation and action we have modeled the proximal behavioral mechanism underling the choices of rhesus monkeys in a simple task that requires them to forage for resources in a dynamic environment. The resulting model suggests that our monkeys have learned to maximize their foraging efficiency given the underlying statistics of reward availability in this task. Moreover, the hidden variables revealed by the model provide us with a framework with which to interpret neurophysiological and brain imaging data collected while monkeys perform the task and to isolate the specific contributions of different brain areas to value based decision making.



Maturation and Plasticity

T17 May 31 - June 3, 2006

MISSED SIGHTS: CONSEQUENCES FOR VISUAL DEVELOPMENT Daphne Maurer, McMaster University Newborns can see but it takes many years for vision to reach adult levels. We have evaluated the contribution of early visual experience to the later development by comparing visually normal children to children who had been deprived of patterned visual input during the first 2-9 months after birth because they were born with dense cataracts in one or both eyes. Longitudinal studies indicate that some aspects of low-level vision normalize after treatment by improving faster than normal to make up for an initial deficit (e.g., contrast sensitivity for low spatial frequencies). For other aspects there are permanent deficits because development asymptotes at a level below normal (e.g., contrast sensitivity for mid-and high-spatial frequencies after binocular deprivation) or because earlier gains are lost (after monocular deprivation). Surprisingly, plasticity for higher-level visual functions cannot be predicted accurately from the results for low-level vision. This point will be illustrated by results from global form, holistic face processing, and biological motion.




T18 May 31 - June 3, 2006

MAPS AND READING DEVELOPMENT IN VISUAL CORTEX Brian Wandell, Stanford University Visual cortex has been an excellent model system for developing a quantitative understanding of brain function. We understand a great deal about the physical signals that initiate vision, and this knowledge has led to a relatively advanced understanding of the organization of major structures in visual cortex, such as visual field maps. This talk will explain several measurements and computational methods that are used to understand human brain development and plasticity. First, we have developed functional magnetic resonance imaging (fMRI) methods for measuring and quantifying the properties of maps in individual human and macaque brains. To understand the development and plasticity of these maps, we have made measurements in several cases of abnormal development as well as in controlled experiments using macaque. Second, we are combining fMRI with diffusion tensor imaging (DTI), a method that can be used to study the white matter fibers, to understand visual development. Specifically, as children develop and learn to read certain visual recognition skills become highly automatized and the brain develops specialized visual circuitry to support skilled reading. We are measuring how certain parts of these circuits develop, and how the signals from these circuits are transmitted to other cortical systems.




T19 May 31 - June 3, 2006

STATISTICAL LANGUAGE LEARNING: COMPUTATIONAL AND MATURATIONAL CONSTRAINTS Elissa Newport, University of Rochester In collaboration with Richard Aslin, I have been developing an approach to language acquisition known as 'statistical language learning.' Our basic idea is that important parts of human language acquisition involve computing, over a stream of speech, such things as how frequently sounds co-occur; how frequently words occur in similar contexts; and the like. The learner then uses these computations to determine regular versus accidental properties of the language being acquired. Our studies have shown that adults, infants, and even nonhuman primates are capable of performing such computations online and with remarkable speed, on both speech and nonspeech materials. However, when tested on more complex computations involving non-adjacent sounds, humans show strong selectivities (they can perform certain computations, but fail at others), corresponding to the patterns which natural languages do and do not exhibit. Primates are not capable of performing some of these more difficult computations.__In addition to this basic statistical mechanism, recent research has revealed that there are maturational changes in the ways in which various types of statistical arrays are compiled into generalizations. Given most types of linguistic input, adults will reproduce the statistics of the corpus that they hear. In contrast, young children will sharpen the statistics, often producing a dramatically more systematic and regular language than the one to which they are exposed. These sharpening processes are also an important part of statistical learning, potentially explaining not only why children acquire language (and other patterns) more effectively than adults, but also how systematic languages may emerge in communities where usages are varied and inconsistent.




T20 May 31 - June 3, 2006

IS MUTUAL INFORMATION THE LEARNING-RELEVANT PARAMETER OF CONDITIONING PROTOCOLS? Randy Gallistel, Rutgers University Two independent lines of evidence indicate that the number of reinforced presentations of a warning stimulus (CS) is, in and of itself, an irrelevant parameter in basic learning protocols, a finding that is, we believe, devastating to all extant associative theories of learning. It is an irrelevant parameter because, for a fixed warning interval (CS-US) there is a perfect trade-off between the number of reinforced presentations required to bring the subject to the point where it makes a conditioned response to the CS and the ratio of the US-US interval to the CS-US interval. If you delete half the trials in a protocol without moving the location in time of the undeleted trials, you double the US-US interval. This operation on protocols has no effect on the progress of conditioning when that progress is plotted as a function of the duration of exposure to the conditioning protocol. The halving of the number of reinforced trials within a given exposure duration is exactly compensated by the doubling of the ratio between the mean US-US interval and the CS-US interval. So what is a relevant parameter. Under a particularly simple-minded calculation of the mutual information between the timing of CS onset and the timing of the US (an analysis that neglects several components of the mutual information), the mutual information between the CS and the US is the relevant parameter of the protocol. Indeed, the assumption that the mutual information unique to a given CS in a given protocol determines not only whether a response will ever develop to a CS but also how long the protocol will have to run for that to happen appears to predict a very wide range of findings in basic conditioning. Thus, the (simplified) unique mutual information between CS and US appears to determine the strength of a protocol, and the duration of exposure to that protocol required before responding begin appears to be inversely proportional to the strength of the protocol. If the rate at which predictive information is acquired is proportional to the strength of a protocol, then this may imply that responding begins when a critical quantity of predictive information has been extracted from exposure to the protocol.

POSTERS

25th Symposium: Poster Abstracts

Center for Visual Science P1 May 31 - June 3, 2006

QUANTIFICATION OF THE VISUAL SYSTEM'S RELIANCE ON A NEWLY RECRUITED CUE USING CUE CONFLICT STIMULI Benjamin Backus, Psychology Dept., University of Pennsylvania Qi Haijiang, Bioengineering Dept., University of Pennsylvania Recent “cue recruitment” experiments used classical conditioning to demonstrate that the visual system can learn to use a new signal, so it becomes a cue that affects the appearance of a bistable stimulus (Haijiang et al., 2006). Pre-existing cues often trade against each other when put into conflict. Here we show that a newly learned cue (stimulus location) trades against a pre-established cue (binocular disparity) in controlling a Necker cube's apparent 3D rotation direction. A plausible interpretation is that visual system assumes conditional independence when first learning a new cue. We quantify the system's reliance on the new cue using d-prime as a measure (Fine & Jacobs, 2002), point out the need to distinguish reliance from reliability. We propose theory to extend weak fusion (weighted averaging of cues) to the degenerate case of bistable percepts within a Bayesian framework, by starting with the beta distribution as the conjugate prior to the binomial distribution. We also show that the new position cue was primarily retinal position (not world position). Fine, I., & Jacobs, R. A. (2002). Comparing perceptual learning tasks: a review. J Vis, 2, 190-203. Haijiang, Q., Saunders, J. A., Stone, R. W., & Backus, B. T. (2006). Demonstration of cue recruitment: Change in visual appearance by means of Pavlovian conditioning. Proc Natl Acad Sci U S A, 103, 483-486. Supported by NIH Grant No. R01 EY 013988.



BAYESIAN INFERENCE WITH PROBABILISTIC POPULATION CODES Jeffrey Beck, Brain and Cognitive Sciences, University of Rochester Weiji Ma, Brain and Cognitive Sciences, University of Rochester Alex Pouget, Brain and Cognitive Sciences, University of Rochester Peter Latham, Gatsby Computational Neuroscience Unit, London Many experiments have shown that human behavior is nearly Bayes optimal in a variety of tasks. This implies that neural activity is capable of representing both the value and uncertainty of a stimulus, if not an entire probability distribution. Here, we argue that the observed variability in neural activity is ideally suited for the representation of the uncertainty. Specifically, we note that Bayes' rule implies that a variable pattern of activity is, in fact, a natural, implicit representation of a probability distribution for the value of a stimulus. Of course, it is by no means clear that the various operations which cortical circuits may perform are capable of manipulating, combining, or decoding such a representation efficiently or otherwise. We address this issue through the construction of a Probabilistic Population Code, or PPC. This type of code consists of two elements: a neural operation, in which the activities of two or more populations of neurons are combined according to biologically plausible rule, and an associated operation on the probability distributions of the variables represented in those population codes, which are obtained through an application of Bayes' rule. Specifically, we show that a neural network which is capable of performing both linear operations and divisive normalization can optimally implement a Kalman filter, but only when the variability of the network units is Poisson-like, that is, when tuning curves are observed and the covariance matrix is proportional to the mean activity. Since both tuning curves and Poisson-like variability are ubiquitous in cortex we propose that a PPC may be a significant mechanism for optimal Bayesian computation.



USING MEMORIZED AND VISUAL LOCATION INFORMATION IN GOAL DIRECTED MOVEMENTS Anne-Marie Brouwer, Center for Visual Science, University of Rochester David C. Knill, Center for Visual Science, University of Rochester People can move their hands to a previously seen target with closed eyes by using the target's memorized location. Is this information also used when visual information about the target is available? Using a virtual display, subjects sequentially picked up and moved two different "magnetic" objects from a target region to a virtual trash bin with their index fingers. In one third of the trials, we perturbed the position of the second object by one centimeter while the finger was transporting the first object to the trash. Subjects did not notice the perturbation. Although the second object was always visible in the periphery, subjects' movements were biased to its initial (remembered) position. The first part of subjects' movements was predictable from a weighted sum of the visible and remembered target positions, with weights of respectively .84 and .16. After approximately 60% of the movement the contribution of memory started to decrease from .16 to .09. Diminishing the contrast of the objects to make them less visible in the periphery doubled the weight that subjects gave to the remembered location. Thus, remembered object location is used to plan goal-directed movements. Reliance on memory increases when the quality of visual information decreases.



THE INFLUENCE OF PRIOR EXPERIENCE ON SACCADE CHOICE Leanne Chukoskie, Salk Institute for Biological Studies Odelia Schwartz, Salk Institute for Biological Studies & HHMI Terry Sejnowski, Salk Institute for Biological Studies & HHMI Richard Krauzlis, Salk Institute for Biological Studies Visually-guided saccades bring items of interest onto the fovea, and have been the subject of intensive study. Under uncertain visual conditions (e.g., fog, dark, or lack of visual structure), eye movements are guided not only by what is observable in the visual world, but also by prior experience about which locations are likely to provide information or reward. However, little is known about the planning of saccades that are guided by non-visual representations of where to look next.We designed a task in which human and macaque monkey subjects searched for a target on a structured noise background (1/f, pink noise). The location of the target in each case did not correspond to any visual element on the screen, but was drawn from a gaussian probability distribution with a given center and spread. Subjects were asked to find the target as quickly as possible; an eye movement to the correct location was rewarded with a tone. Subjects' eye movements revealed that they had learned both the location of the center of the probability distribution from which the targets were drawn and also had some knowledge of the shape of the distribution. In addition, subjects seemed to utilize the visual landmarks in the pink noise background despite the fact that they were uncorrelated with target location.We conclude that humans were able to build estimates of where to look even when visual cues do not provide information about the location of the rewarded target. In addition, our new task provides a framework for probing how prior information is integrated with visual information to decide where to look.



INTEGRATING VISUAL INFORMATION INTO PROBABLISTIC MODELS OF WORD LEARNING Michael C. Frank, Brain and Cognitive Sciences, MIT Vikash Mansinghka, Brain and Cognitive Sciences, MIT Edward Gibson, Brain and Cognitive Sciences, MIT Infants have the ability to learn distributional information in a variety of sensory modalities (Aslin, Saffran, & Newport, 1998; Fiser & Aslin, 2002; Saffran, Johnson, Aslin, & Newport, 1999). We asked whether adults are able to use distributional information from two modalities simultaneously in the service of learning the form and meaning of words in a simple artificial language. Participants in our experiment heard randomly generated sentences of unsegmented, synthesized speech created via concatenation of words of various lengths. Each sentence contained both a random number of filler words and exactly one meaning word which corresponded to a simultaneously presented picture of a novel object, all arranged in a random order. A control condition was identical save that the association between the meaning words and the novel objects was not fixed, so no meaning word was associated with any particular object. Participants were tested both on the forms of the words they heard (as in Saffran, Aslin, & Newport, 1996) as well as on the correspondence between particular meaning words and objects. Participants in the experimental condition learned both word meanings and word forms at a level significantly greater than those in the control condition (t(568) = 3.68, p < .001, and t(1141) = 4.70, p < .001, respectively), suggesting that they were able to use distributional information from two different modalities to associate words in unsegmented speech with their referents. We present a word-based computational model of segmentation and meaning-learning—derived from that of Brent (1999)—which accounts for these results Aslin, R. N., Saffran, J. R., & Newport, E. L. (1998). Computation of conditional probability statistics by 8-month old infants. Psychological Science, 9, 321-325. Brent, M. R. (1999). An efficient, probabilistically sound algorithm for segmentation and word discovery. Machine Learning, 34, 71-105. Fiser, J. & Aslin, R. N. (2002). Statistical learning of new visual feature combinations by infants. Proceedings of the National Academy of Sciences, 99, 15822-15826. Saffran, J. R., Johnson, E. K, Aslin, R. N., & Newport, E. L. (1999). Statistical learning of tone sequences by human adults and infants. Cognition, 70, 27–52. Saffran, J.R., Newport, E.L., & Aslin, R.N. (1996). Word segmentation: The role of distributional cues. Journal of Memory and Language, 35, 606-621.



ABILITY TO TASK-SWITCH IN ACTION VIDEO GAME PLAYERS C.Shawn Green, Brain and Cognitive Sciences, University of Rochester Daphne Bavelier, Brain and Cognitive Sciences, University of Rochester The ability to switch between competing goals depending on task demands is a key determinant of cognitive skills. This work investigates whether action video game players (VGPs) show changes in this ability. We first compared two versions of the attentional blink task, one with a task-switch between T1 and T2 and one without. Previous work documents a larger blink in the task-switch than in the non-task-switch condition, indicating an added cost of task switching (Potter et al., 1998). Accordingly, the size of the blink was greater in the task-switch version (identify T1/detect presence of T2) than in the non-task-switch version (identify T1/identify T2) in non-video game players (NVGPs). However, in addition to displaying overall reductions in blink depth compared to NVGPs, VGP performance was equivalent with and without a task-switch, suggesting a lesser cost of task-switching in VGPs. To directly assess task-switching ability, a standard task-switching paradigm was employed in which subjects were presented with colored shapes (red/blue circle/square) and were asked to predictably switch between naming the color and the shape of the item. While both groups showed the typical task-switch cost (a large increase in reaction time on the switch trials), this increase was significantly smaller in VGPs than in NVGPs, establishing enhanced task-switching abilities in VGPs. Furthermore, the VGP effect was similar whether the subjects used a manual or vocal method of response, suggesting that the results are likely not due to a greater ability of VGPs at performing arbitrary stimulus-response mappings.



TRAINING-INDUCED IMPROVEMENTS IN MOTION SENSITIVITY AFTER V1 DAMAGE IN HUMANS Kristin Kelly, Ophthalmology & Center for Visual Science, University of Rochester Brian Sullivan, Brain and Cognitive Sciences, University of Rochester Mary Hayhoe, Brain and Cognitive Sciences & Center for Visual Science, University of Rochester Krystel Huxlin, Ophthalmology & Center for Visual Science, University of Rochester Damage to the adult primary visual cortex (V1) is a principal cause of blindness in humans. Although some initial, spontaneous recovery may occur due to resolution of inflammation around the lesion site, permanent loss of visual perceptual abilities usually ensues. Visual motion perception, which is critical for navigating and interacting with our dynamic environment, is one of the visual modalities permanently impaired by V1 lesions. However, pilot studies in a cat animal model suggest that intensive motion discrimination training in the blind field can induce localized recovery of motion thresholds following permanent cortical damage. To determine whether such training-induced recovery might be possible in humans, three patients with homonymous visual field defects were recruited one year after a cortical stroke. Following baseline evaluation of visual performance, they were required to perform 300 trials per day of a two-alternative, forced-choice, global direction-discrimination task in their blind field. Global motion signals are normally processed by higher-level visual cortical areas such as the MT+ complex which was intact in all patients. Daily repetition of this task at a single location in the blind field for 3-6 months improved global motion thresholds at that site, enlarged the patients' Humphrey visual fields, and improved detection and tracking of moving objects in a 3D, naturalistic, virtual environment. Thus, it appears that human visual motion perception can be improved following V1 damage using intensive direction discrimination retraining with complex, dynamic stimuli presented in the blind field. The exact cortical mechanisms mediating this recovery are currently under investigation.



ATTENTION CAN ALTER THE TEMPORAL CAPACITY OF OBJECT PROCESSING IN HIGH-LEVEL VISUAL AREAS Thomas J. McKeeff, Department of Psychology, Princeton University Frank Tong, Department of Psychology, Vanderbilt University Attention has an important role in our ability to individuate objects across variations in both space and time. Many studies have investigated how attention can dynamically alter the spatial tuning properties of visual neurons, but much less is known about whether attention can alter the temporal properties of the visual system. Previously, we have shown that temporal tuning functions of individual human visual areas can be reliably measured with fMRI. Here, we investigated whether spatial attention can enhance the temporal processing capacity of cortical visual areas during object processing. Subjects were instructed to attend to one of two simultaneously presented RSVP sequences of face and houses, which were presented to the left and right of central fixation. Presentation rate varied from 2-30 items/second. Spatial attention led to an overall increase in response amplitudes in early areas V1-V3, but did not alter the temporal frequency response profile of these areas. In V4v and the fusiform face area, attention not only led to enhanced response amplitudes but also led to a rightward shift in the temporal frequency response profile, indicating a shift in peak sensitivity toward higher temporal rates. In comparison, the parahippocampal place area showed weak evidence of attentional modulation. Our results suggest that spatial attention is capable of altering the temporal processing capacity of some high-level visual areas. These results may be of functional significance when an observer must identify an object under temporally demanding conditions.



DOES PERCEPTION LEARNING ACQUIRE THE PRIOR? Monica Padilla, Department of Biomedical Engineering and Center for Vision Science and Technology, University of Southern California Norberto M. Grzywacz, Department of Biomedical Engineering, Center for Vision Science and Technology and Neuroscience Graduate Program, University of Southern California Models of Statistical Learning Theory (Vapnik, 1999) use examples of the input and of supervisor answers to optimize the performance of tasks. Unfortunately, although powerful, these models may not be directly applicable to the brain. Brain centers typically have no direct access to the relevant input but to noisy, nonlinearly processed versions of it. To remediate this problem, we have proposed a Bayesian learning framework, which is a generalization of Statistical Learning Theory (Grzywacz and Padilla, 2006). This generalization allows learning of all underlying functions of Bayesian Decision Theory. In particular, a system can learn the prior distribution, the likelihood function, or the decision rule. The most powerful of these functions is the prior. With it, a system can perform internal experiments and optimize the other two functions for any task. Here, we use this prediction to test whether perceptual learning can acquire the prior distribution. We trained subjects in speed-segmentation learning tasks. A display of moving dots could appear with the faster speed on the left (1) or on the right (2), or with the fast and slow dots intermixed (3). Each experimental condition had a defined prior (i.e., probability of each of these three possibilities). Moreover, each condition had a specified task (e.g., determining which side was faster or determining whether the display had segmentation). Subjects showed a typical learning curve when consecutive conditions had different tasks but same priors. Hence, subjects either cannot learn the prior information or do not use it efficiently when performing new tasks. The work was supported by National Eye Institute Grants EY08921.



HOMONYMOUS HEMIANOPIA ALTERS THE DISTRIBUTION OF FIXATIONS IN 3D VIRTUAL ENVIRONMENT Meghan Riley, Ophthalmology, University of Rochester Tim Martin, Ophthalmology & Center for Visual Science, University of Rochester Kristin Kelly, Ophthalmology & Center for Visual Science, University of Rochester Mary Hayhoe, Brain and Cognitive Sciences & Center for Visual Science, University of Rochester Krystel R. Huxlin, Ophthalmology & Center for Visual Science, University of Rochester Homonymous hemianopia is a severe form of cortical blindness that impairs visual function in everyday life. Hemianopic fixation patterns are abnormal during visual search and during examination of 2D images. Our goal was to compare eye movements in 4 individuals with homonymous hemianopia and 4 visually normal subjects. During immersion in a naturalistic, 3D virtual environment, eye positions were recorded under 3 conditions: (1) while stationary and fixating a point in the environment, (2) while stationary but freely scanning the environment, (3) while walking an L-shaped path while freely-scanning the environment. As previously reported for 2D images, hemianopic subjects placed in a 3D virtual environment fixated significantly more into the blind versus intact halves of their virtual field of view, regardless of experimental condition. During walking, hemianopic fixations were 1.5x more narrowly distributed and 24% more lateralized than in walking normals. Hemianopic fixation durations hovered between 300 and 600ms regardless of condition. Normal fixation durations averaged 1100ms when required to fixate a static point in the environment, but decreased to ~350ms during both free-gazing conditions, suggesting possible differences in fixation dynamics and quality of visual information extracted between hemianopes and normals during fixation of pre-determined targets. Overall, compensatory hemianopic eye movement strategies used for static 2D images generalized to more complex, 3D environments. However, additional deviations from normal visual behavior emerged that were modulated by the type of interaction with the virtual environment. Understanding these deviations could provide new means of studying the visual strategies developed by hemianopes, and of assessing their consequences for visual function in every day life.



ACTION VALUE IN THE SUPPLEMENTARY AND PRESUPPLEMENTARY MOTOR AREAS Jeong-Woo Sohn, Brain and Cognitive Sciences, University of Rochester Bruno B. Averbeck, Brain and Cognitive Sciences, University of Rochester Daeyeol Lee, Brain and Cognitive Sciences, University of Rochester The mapping between a particular action and its outcome is often stochastic and it may be revealed after a variable delay, which makes it difficult to find an optimal action-selection rule. An efficient solution to this problem is provided by reinforcement learning algorithms in which an ongoing estimate for the temporally discounted sum of future rewards, known as the value function, is used as the basis for choosing optimal actions. In this study, we investigated whether and how the neurons in the supplementary and pre-supplementary motor areas (SMA and pre-SMA) encode action values. Two rhesus monkeys were trained to produce a series of hand movements by manipulating a joystick in response to visual targets. The sequence of target locations and the location of the rewarded target were fixed in a given block of trials, providing the animal with the information necessary to determine the number of remaining movements (NRM) before reward. In addition, the position of the first target in a given trial was varied randomly within a block of trials, in order to manipulate separately the NRM and the ordinal position of each movement. Single-unit activity was recorded from 110 SMA neurons and 114 pre-SMA neurons, and the activity during a 500 ms interval beginning 250 ms before target onset was analyzed separately according to whether the NRM for a given movement was known before target onset or whether the NRM was ambiguous due to the uncertainty in the direction of upcoming movement. First, we found that the responses of individual neurons often changed according to the action values of individual movements, but they were also influenced by movement directions. To determine how the effects of the action values and movement directions were combined, we factored out movement-related activity with a generalized linear model and compared the performance of additive and multiplicative models with cross-validation. The results showed that in the majority of neurons in both SMA and pre-SMA, signals related to action value and movement direction were combined in a multiplicative manner. This suggests that action value changes the gain of movement-related signals. Second, we found that when there was uncertainty in the directions of upcoming movements, anticipatory activity in SMA and pre-SMA tended to encode the direction of the movement with the higher action value. For example, when the direction of the next movement was uncertain, neurons tuned for right-ward movements tended to increase (decrease) their activity if the rightward movement was associated with a higher (lower) action value. By applying a linear discriminant analysis with a sliding window, we confirmed that when the direction of the next movement was uncertain, neural activity tended to encode initially directions of movements with higher action values, but, after target onset, quickly switched their activity according to the executed movement. In summary, these results suggest that the medial frontal cortex plays a key role in the ongoing evaluation of action values for alternative movements.



ACQUIRING AND PROCESSING VERB ARGUMENT STRUCTURES: A MINIATURE LANGUAGE STUDY Elizabeth Wonnacott, Brain and Cognitive Sciences, University of Rochester Elissa L. Newport, Brain and Cognitive Sciences, University of Rochester Michael K. Tanenhaus, Brain and Cognitive Sciences, University of Rochester Adult language combines a complex mix of regular, ‘rule like’ processes and more conservative, lexically based patterns. For example, verb argument structure constructions may generalize to new verbs (John gorped ->Bill gorped John) yet resist generalization with certain lexical items (John sighed -> *Bill sighed John). It has been suggested that that whether a particular verb can occur with a particular argument structure may depend upon a fine-grained semantic representation (e.g. Pinker 1989). However, more recent work suggests that statistical learning mechanisms play an important role in acquisition (e.g. Saffran, Aslin and Newport 1996), and frequency-based entrenchment effects in young children (e.g. Theakston 2004) indicate that verb argument structure acquisition may closely depend upon the distribution of forms in the input. Moreover, on-line comprehension demonstrates sensitivity to the statistical likelihood of a particular verb occurring with a particular construction (e.g. MacDonald et al 1994). The current research program investigates whether learners exposed to Miniature Languages track verb-structure co-occurrences, as well as the distribution of argument structures across the language. We explore how this statistical information becomes reflected in three different language behaviors: production, grammaticality judgment, and on-line sentence processing.



THE DEVELOPMENT OF OBJECT AND FACE PROCESSING IN CHILDREN Jennifer Yoon (Davie), Psychology Department, Stanford University Kalanit Grill-Spector, Psychology Department, Stanford University Previous research has indicated that high-level vision in adults follows a surprising sequence of processing steps, with identification (e.g., Harry vs. George) following categorization (e.g. face vs. car), but with categorization behaviorally indistinguishable from detection ( e.g. object vs. nonobject)1. Here we asked whether object and face perception during childhood involves developmental changes in the sequence of processing stages found in adults, and/or developmental changes that are domain-specific, following a different trajectory for faces versus non-face objects. To address this question, we tested children ages 5 through 17, as well as adults (> 18 years) in a series of psychophysical experiments in which naturalistic photographs were presented for brief exposure durations (parametrically varied: 17, 33, 50, 67, 150ms), then masked with a scrambled image. Participants made forced-choice judgments about each photograph. In different blocks, judgments varied by processing step (detection, categorization, and identification) and by domain (faces, cars). For example, in face identification task, participants viewed photographs of children, deciding whether the child is or is not Harry Potter. In the car identification task, participants viewed images of cars and judged whether the car is or is not a Jeep. Preliminary results from adults show that both processing stages and performance for faces and non-face objects (cars) are similar. Preliminary data on 9- to13-year-old children show that, similar to adults, their performance was better for categorization than identification of both cars and faces. Further results will determine whether this processing sequence undergoes developmental changes after 5 years of age, and whether these changes follow distinct trajectories for different domains of stimuli (e.g faces vs. objects). Grill-Spector, K. & Kanwisher, N. (2005). Visual recognition: As soon as you know it is there, you know what it is. Psychological Science, 16(2), 152-160.

MEETING ATTENDEES

Rebecca Achtman, University of Rochester, [email protected]

Alison Austin, University of Rochester, [email protected] Bruno Averbeck, University of Rochester, [email protected]

Benjamin Backus, University of Pennsylvania, [email protected] Galina Badyulina, University of Rochester, [email protected]

Neil Bardhan, University of Rochester, [email protected]

Dominic Barraclough, CVS University of Rochester, [email protected] Daphne Bavelier, U. of Rochester, [email protected]

Jeffrey Beck, University of Rochester, [email protected] Anne-Marie Brouwer, Center for Visual Science, University of Rochester,

[email protected]

Ellen Campana, University of Rochester, [email protected] Aaron Cecala, University of Rochester, [email protected]

Kyle Chambers, University of Rochester, [email protected]

Nick Chater, Psychology, University College London, [email protected] Leanne Chukoskie, Salk Institute for Biological Studies, [email protected]

Marvin Chun, Yale University, [email protected] Meghan Clayards, University of Rochester, [email protected]

Nathaniel Daw, Gatsby Computational Neuroscience Unit, [email protected]

Elizabeth DeGrush, Brain and Cognitive Sciences, [email protected] Peter Delahunt, Posit Science Inc., [email protected]

Duane Dey, Bausch and Lomb, [email protected] Charles Duffy, Univ of Rochester, [email protected]

Michael Frank, MIT, [email protected]

Austin Frank, University of Rochester, [email protected] Charles Gallistel, Rutgers University, [email protected]

Margaret Gardner, University of Rochester-BCS, [email protected] Andrea Gebhart, University of Rochester, [email protected]

Bernard Gee, University of Rochester-Neuroscience, [email protected]

Krista Gigone, Center for Visual Science, University of Rochester, [email protected]

Deda Gillespie, McMaster University, [email protected]

Jason Gold, Indiana University, Bloomington, [email protected] Daniel Goldreich, McMaster University, [email protected]

C. Shawn Green, University of Rochester, [email protected] Hal Greenwald, University of Rochester, [email protected]

Arnaud Guidon, Rochester Center for Brain Imaging-University of Rochester,

[email protected] Peter Haake, Summer Fellowship , [email protected]

Andrew Herbert, Psychology, RIT, [email protected] Nicholas Hindy, Cornell University, [email protected]

Elizabeth Hirshorn, University of Rochester, [email protected]

Lori Holt, Carnegie Mellon University, [email protected] Bo Hu, University of Rochester, [email protected]

Vincent Huang, Johns Hopkins University, [email protected]

Krystel Huxlin, University of Rochester Eye Institute, [email protected] Jaewon Hwang, Center for Visual Science, [email protected]

Seo Hyojung, University of Rochester, [email protected] Robert Jacobs, University of Rochester, [email protected]

Daniel Johnson, University of Texas at San Antonio, [email protected]

Kristin Kelly, University of Rochester, [email protected] Soyoun Kim, University of Rochester, [email protected]

Natalie Klein, University of Rochester, [email protected] Talia Konkle, MIT, [email protected]

Sarah Laredo, Brandeis University, [email protected]

Terri Lewis, McMaster University, [email protected] Renjie Li, University of Rochester, [email protected]

Wei Ji Ma, University of Rochester, [email protected] Anthony Maida, Univ Louisiana at Lafayette, [email protected]

Walter Makous, Center for Visual Science, [email protected]

Daniel Margoliash, University of Chicago, [email protected] Tim Martin, Ophthalmology, CVS, [email protected]

Osamu Masuda, CVS, Univ. of Rochester, [email protected]

Nestor Matthews, Denison University, [email protected] Daphne Maurer, McMaster University, [email protected]

Thomas McKeeff, Princeton University, Department of Psychology, [email protected]

William Merigan, Rochester, [email protected]

Ross Messing, University of Rochester, [email protected] Melchi Michel, University of Rochester, [email protected]

Toby Mintz, University of Southern California, [email protected] Richard Murray, York University, [email protected]

Elissa Newport, University of Rochester, [email protected]

Tomokazu Oshiro, University of Rochester, [email protected] Monica Padilla, University of Southern California, [email protected]

William Page, University of Rochester, [email protected]

Gary Paige, Univ. of Rochester, [email protected] Tatiana Pasternak, Center for Visual Science, [email protected]

Alexandre Pouget, University of Rochester, [email protected] Vikranth Rao, University of Rochester, [email protected]

Rachel Robbins, McMaster University, [email protected]

Jeff Saunders, University of Pennsylvania, [email protected] Keith Schneider, Rochester Center for Brain Imaging, [email protected]

Reza Shadmehr, Johns Hopkins University, [email protected] Aimee Slaughter, University of Rochester, [email protected]

Jeong-Woo Sohn, University of Rochester, [email protected]

Rebecca StClair, University of Chicago, [email protected] Leo Sugrue, HHMI/Stanford University, [email protected]

Brian Sullivan, University of Rochester - Center for Visual Science, [email protected]

Mario Svirsky, NYU School of Medicine, [email protected]

Duje Tadin, Vanderbilt University, [email protected] Michael Tanenhaus, University of Rochester, [email protected]

Stephen Teitel, UR Physics, [email protected]

Josh Tenenbaum, MIT, [email protected] Chia-huei Tseng, University of California, Irvine, [email protected]

Nicholas Turk-Browne, Yale University, [email protected] Jennifer Vannest, Brain & Cog Sci, University of Rochester,

[email protected]

Ed Vul, MIT, [email protected] Brian Wandell, Stanford University, [email protected]

David R Williams, Univ. of Rochester, Center for Visual Science, [email protected]

Uta Wolfe, Hobart and William Smith Colleges, [email protected]

Elizabeth Wonnacott, University of Rochester, [email protected] Jennifer Yoon, Stanford University, [email protected]

Anthony Zador, Cold Spring Harbor Laboratory, [email protected]

25th Symposium of the University of Rochester Rochester ... › media › pdf › symposium ›...

Documents

Transcript of 25th Symposium of the University of Rochester Rochester ... › media › pdf › symposium ›...