Jeremy Wolfe
Visual Search: Is it a matter of life and death?
Like Gaul, this talk is divided into three parts:
1) I will give an introduction to the problem of visual search and to the Guided Search model that my
lab has been working on for a number of years.
2) I will discuss an important practical problem in search. Rare targets are hard to find simply
because they are rare. We ask people to find rare targets in some very important tasks like baggage
screening and routine mammography so, if low target prevalence makes search difficult, this could be
a real problem.
Finally, I will place the problem of search into the larger context of visual perception and show how
our need to use selective attention leads to some interesting perceptual errors.
|
|
|
Anina Rich
Is attention critical in synaesthesia? Review and new data
Synaesthesia is a fascinating phenomenon in which particular stimuli elicit unusual involuntary experiences. For example, a
sound may induce a colour experience. In the most common form, seeing a letter results in a colour that is vivid and highly
consistent over time. Synaesthetes get interference from these colours in Stroop-type tasks, but there is considerable debate
regarding the necessity for inducing stimuli to be attended in order for synaesthesia to occur. I will review the evidence for
and against 'preattentive' theories of synaesthesia and present data from a modified attentional blink task that suggest an
important role for attention in this unusual phenomenon.
|
|
|
Tania Lombrozo
Simplicity and probability in explanation
Philosophers and scientists have extolled the virtues of simplicity for
centuries. William of Occam, for example, is known for his stricture not
to
invoke more entities in explanation when less will do. Not surprisingly,
contemporary undergraduates share the preference for simpler explanations.
But the simplest explanation is not always the most likely. How is the
preference for simpler explanations influenced by probability information?
In this talk, I'll explore two related questions. First, how do simplicity
and probability trade-off in evaluating explanations? And second, do
explanatory criteria like simplicity influence probability judgments? That
is, do people believe simpler explanations are more likely to be true? The
data I'll report suggest that both simplicity and probability influence
how
explanations are evaluated, but that disproportionate probability
information is required to override the preference for simpler
explanations.
What's more, committing to a simple but unlikely explanation can distort
participants' estimates of frequency information. These findings are used
to
motivate the idea that the process of generating and evaluating
explanations
plays an important role in assessing subjective probability.
|
|
|
Sam Moulton
Does psi exist?
In its most conservative instantiation, the psi hypothesis asserts
the anomalous presence of knowledge. A direct and promising strategy
to quantify potential psi effects lies in neuroimaging approaches
that make few assumptions about the nature of such knowledge. By
circumventing behavior and directly examining brain activity we can
increase the likelihood of uncovering a psi effect, and decrease the
ambiguity of a null result. In this study, we relied upon the robust
finding within cognitive neuroscience that prior knowledge of a
stimulus biases the brain?s response to that stimulus, even when the
prior knowledge is unconscious. We hypothesized that knowledge
acquired through psi would similarly prime neuronal activity, and
that evidence of priming associated with psi stimuli would offer
compelling evidence for the existence of psi. Sixteen participants
completed a simple binary guessing task while their brain activity
was monitored using functional magnetic resonance imaging (fMRI) . In
this task, participants viewed two sequential photographs and were
asked to decide which image would shortly be randomly chosen by the
computer; after participants indicated their choice on each trial,
they were shown the target stimulus from the pair a second time.
Outside the scanner, an emotionally-close or biologically-related
partner viewed only the target stimuli. Thus all three hypothesized
mechanisms of anomalous information acquisition -- precognition,
telepathy, and clairvoyance -- could affect performance in this
design. We analyzed Blood Oxygen Level-Dependent (BOLD) activity
associated with target versus decoy stimuli in search for evidence of
psi.
|
|
|
Matthew Finkbeiner
Domain-specific processing in the left fusiform gyrus: Evidence from
distractor interference and extinction
Recent neuroimaging studies have revealed that letters activate both the
left and right fusiform areas, but that only the left fusiform responds
to letters more than to control stimuli. Though these findings suggest
that the left fusiform is specialized in its function of identifying
letters, it does not rule out the possibility that the right fusiform
contributes critically to letter identification processes. To test this
possibility, we used a behavioral word identification task with
bilateral and unilateral displays to determine the cost of engaging the
right hemisphere with a distractor stimulus. We found that while
engaging the left hemisphere led to a robust interference effect,
engaging the right hemisphere had no effect at all. In a separate
neuroimaging study, participants were asked to indicate the presence of
a flicker in rapidly presented streams of letters. Despite letters not
being relevant to the task, fMRI revealed a robust modulation by
attentional capture in the left (but not right) fusiform gyrus. Taken
together, these findings indicate that the left fusiform is tuned to
process letters and, hence, becomes selectively engaged (captured)
whenever attention is drawn to the presence of letters.
|
|
|
Matt Bronstad
Understanding facial beauty: A computational model of attractiveness
judgements
Humans share an aesthetic standard for faces, as people from different
cultural and economic backgrounds tend to agree about which faces are
attractive. Much research in the last two decades addressed the question
of which aspects of facial appearance people use in forming
attractiveness judgments. In particular, researchers sought to determine
whether people are primarily sensitive to facial averageness, symmetry,
or sexual dimorphism. We still have little understanding of how different
aspects of appearance codetermine perceptions of attractiveness. I will
present a computational model that replicates human attractiveness
judgments given images of faces and attractiveness ratings of those
images. The ratings and images are reduced into simpler image factors
that are sensitive to aspects of facial appearance. These factors are
strikingly similar to averageness, symmetry, and sexual dimorphism.
|
|
|
Elissa Aminoff
Contextual associations: Implications for object recognition and memory
Objects in our environment are clustered in typical contexts.
Consequently, when you walk into a room and see a bed, for example, not
only do you immediately identify the room as a bedroom, but you also
expect to find nearby specific objects with high likelihood (e.g., a night
table, an alarm clock, a dresser). This clustering of objects by
contextual associations, can be used to increase efficiency in processing
the wealth of perceptual information in our environment. Using fMRI, we
have demonstrated that such contextual associations are processed in the
parahippocampal cortex and the retrosplenial complex. This contextual
processing, which usually facilitates perception and cognition, can
sometimes elicit distortions of memory. Specifically, we hypothesized that
seeing a highly contextual object (e.g., a desk) would result in
contextual co-activation such that observers might later falsely perceive
a related object (e.g., an office chair) as having been previously
presented. Furthermore, would brain activity in the context cortical
network be more active at the time of encoding (i.e. viewing the desk)
because of the increased activation of contextual associations when a
false memory is likely to occur? We studied how contextual associations
can affect memory and what brain areas predict the occurrence of a false
memory. Results demonstrated that subjects are more likely to false alarm
to contextually related item than to unrelated novel items. fMRI results
showed differential activity that correlated with subsequent false
recognition in the retrosplenial complex, but not in the parahippocampal
cortex. The results from this study will be discussed in a larger
framework of the role of the cortical context network and the influence of
contextual associations on memory and its distortion.
|
|
|
Joo-Hyun Song
Characteristics of visually-guided manual pointing: Continuous interaction
between cognitive decision-making & actions
We examined characteristics of visually-guided manual pointing in humans,
showing continuous interaction between cognitive decision making, and
actions. In addition, results of our study demonstrate advantages of
visually-guided action as a research tool to read out continuous cognitive
processes over time. We showed that manual pointing directed to multiple
stimuli can be concurrently planned and executed before decision-making is
completed. This result demonstrates flexibility in visuo-motor control,
casting doubt on prevalent assumptions regarding the existence of a unique
solution for optimal control. We also demonstrated that the readiness of
motor execution is not simply determined by the current trial itself.
Manual pointing performance for the same trials depends on whether
difficulties of trials are varied or maintained the same within a block.
The result shows that it is not explicit knowledge of upcoming trials but
history of the very recent past, which determines the current state of
performance. Last, we show the applicability of visually- guided pointing
in a number task, to reveal other properties of cognitive processing.
|
|
|
Alvaro Pascual-Leone
What do the Blind teach us about Seeing?
Much of the normal human brain appears to be dedicated to process visual
information and vision provides a major source of information about our
environment and a guide to our actions. What happens to all those parts of
the brain in a person born blind? The human brain is capable of remarkable
change and appears able to keep changing throughout life. These changes
are summarized under the term plasticity and appear to be driven by a
number of factors, including changes in the environment or in the
information that reaches the brain. Therefore, it is reasonable to presume
that in the case of blindness, the parts of the brain normally devoted to
process visual information may change and become involved in the
processing of tactile or auditory information, or even in memory,
language, or other cognitive tasks that might help the blind person cope
better with the challenges of the lack of vision.
Indeed, in studies using positron emission tomography (PET) or functional
magnetic resonance imaging (fMRI) subjects born blind show activation of
their "visual" cortex when reading Braille, performing various other
tactile tasks, identifying the location of a sound, and even completing a
verbal memory task. This activation of the "visual" cortex for
non-visual tasks is correlated with how well the blind subjects perform in
the non-visual tasks, so that the more activation the better the sound
localization ability or the verbal memory. In fact, congenitally blind
subjects are better than sighted subjects at these non-visual tasks,
perhaps because they are able to use the additional resources provided by
the "visual cortex". Consistent with this idea, transient disruption of
the visual cortex in early blind subjects using non-invasive transcranial
magnetic brain stimulation (TMS) significantly impairs their ability to
perform certain tactile, auditory, verbal memory and language tasks.
Therefore, in the early blind, parts of the brain that normally process
visual information are used for other non-visual functions. In this sense,
the study of the blind brain provides a unique opportunity to investigate
the plastic changes that can occur in response to changes in the input of
information to the brain and inquire about their eventual behavioral
consequences. Such issues of plasticity are critical for the adaptation to
blindness and for novel approaches to restore sight to the blind.
Blindness and visual loss are pervasive problems in our society and
important causes of major disability. The insights discussed are of
relevance for rehabilitation and education programs for the blind,
including mobility, Braille learning, spatial orientation and object
representation questions.
However, studying how the brain of blind people processes certain
information provides also an opportunity to learn about visual perception
and the organization of the brain in the sighted. In sighted subjects
wearing a blindfold and thus being completely deprived of vision for even
just a few days, there are dramatic changes in brain activity in response
to tactile, auditory or verbal stimuli, that resemble those encountered in
the blind. This extremely rapid plasticity suggests that non-visual
information may be reaching the "visual cortex" even in normal sighted
humans subjects by connections in the brain that that can be strengthened
quickly in case of loss of vision. In this sense, comparing sighted,
blindfolded, and blind people as they interact with their surrounding
environment, recognize objects, handle tools, walk safely around their
house or their city, locate the source of a sound or identify a friend by
her voice, can teach us fundamental aspects of the workings of the human
brain. Such studies using fMRI and TMS suggest that brain regions that
were thought to be devoted to processing vision, do in fact respond to the
sensorial information of the stimuli (such as object form, spatial
localization, movement direction) independently from the sensory modality
- vision, hearing or touch - that conveys the information to the brain.
This implies that the brain may be organized in areas that do a certain
operation regardless of the sensory modality that conveys the required
information. An important issue for future research is to understand the
normal brain connectivity between regions processing visual, tactile,
auditory, and other cognitive information, and how such connectivity
changes in the visually deprived.
|
|
|
Tim Vickery
Interactions of feedback and attention during simple choice-making
We explored competitive choice behavior and its neural underpinnings
in several
experiments using a simple game. In behavioral experiments, we
examined the ability of
humans to produce 'random' sequences of choices in a game context. We
gave or witheld
feedback in a 'penalty kick' game against a demanding AI opponent,
and found that
subjects produced more random sequences with feedback than without,
although the
sequences produced with feedback were still distinguishable from
randomly generated
sequences. By sometimes applying a taxing secondary task, we also
found that this
ability is dependent on central attentional resources. In fMRI
experiments, we
measured BOLD activity in humans as they played this game while being
scanned. We
devised criteria to identify regions of the brain that are selective
for making
uncertain choices on the basis of feedback, and distinguish these
from regions that
are more active when a task demands more effort or attention.
Applying these
criteria, we classified a region of the right inferior parietal
lobule and a subset of
frontal regions as specifically involved in choice-making under
uncertainty.
|
|
|
John Gabrieli
Enhancing Learning in the Human Brain: Evidence from
Functional Neuroimaging
Most of our knowledge about the organization of learning and memory
in the human brain comes from studies of patients with focal lesions
who exhibit impaired learning (amnesia). Functional neuroimaging now
allows for the study of neural systems that promote human learning.
I will review recently completed studies using functional magnetic
resonance imaging (fMRI) in which we examined how self-knowledge and
motivation can promote declarative or episodic memory formation. I
will also review a study in which we used real-time fMRI as a source
of feedback to allow people learn how to gain direct control of
activation in a targeted brain region, which in turn evoked the
mental functions mediated by that brain region.
|
|
|
Kevin Shapiro
Nouns, Verbs, and the Brain
The grammatical categories 'noun' and 'verb' are universal
across languages and represent the elementary building blocks of
propositional speech. Nevertheless, it isn't clear how the brain
categorizes words as nouns and verbs, or at what stage this information
enters into the process of word production. For example, some have argued
that the noun/verb distinction is essentially a semantic one, capturing a
fundamental difference in meaning between words assigned to the two
categories. Others view the distinction in syntactic terms, pointing out
that nouns and verbs play different grammatical roles and are subject to
different kinds of morphological processes. Using evidence from
neuropsychological patients with impairments in word production, I will
argue that apparently selective deficits in noun or verb processing can
arise at either the semantic or the synactic level, leading to different
patterns of errors in speech and comprehension. I will then describe
several experiments using functional magnetic resonance imaging (fMRI) and
transcranial magnetic stimulation (TMS) that may shed light on the brain
areas that subserve core semantic and syntactic processes in noun and verb
production.
|
|
|
|
