367 Midterm 2

about how many discriminable items to the avg adult?

30,000

What info is required to identify a distal stimulus?

Light source, reflectance, surface orientation, viewing position

What is the inverse position problem?

how do you go from the proximal stimulus to identifying the distal stimulus?

What is underdetermination?

any 2-D image may have many potential 3-D sources that created it

Viewpoint invariance

ability to recognize an object seen from different viewpoints; difficult for computers

Contour

border between light and dark areas in an image

Ganzfield ("total field"):

sort of sensory deprivation; developed to test for telepathic communication

What is does Troxler fading in Ganzfield indicate?

sharp luminance contours must be important to visual system

Microsaccades and their purpose

minute, involuntary, almost imperceptible eye movements that occur during fixations. change the position of the image on the retin

Eye responds best to what type of stimuli?

Dynamic

Basis of Gestalt psychology?

movement could be produced by a succession of stationary stimuli, apparent motion cannot be explained by sensations alone

What does GP minimize the role of?

Past experiences

What does GP oppose and say instead?

structuralism; "the whole is different than the sum of its parts"

Illusory contours

contours not present on retina, but still affect contour perception

what cells are acvtivated by illusory contours?

end-stopped cells

What two laws overpower proximity?

common region and uniform connectedness

Pro of GP

holistic

Cons of GP

operates post hoc; not good for predictions,

not good at explanations,

vague definitions: "simplicity"?

What theories replaced GP?

Template theories; compare input to a model or "template" stored in memory

Pro of template theory

successfully used by machines

cons of template theory?

intolerant to variations in a stimulus, too many templates required, cannot handle novel stimuli, cannot handle context

What type of theory is pandemonium?

a feature-based theory

4 Steps of Pandemonium (Selfridge)

image demon (sensory input), feature demon (analyze features), cognitive demon (WHAT is present), decision demon (identify pattern)

Pros of feature theory

evidence: "feature-detectors" physiologically relate to cells in visual system, can identify wide range of stimuli--just specify component features

Cons of feature theory

doesn't define "feature": single line segment? two lines forming an angle? curved line?, cannot handle Gestalt organizational principles: when is a row of dots a line?

\
, cannot handle context

\
, what about real-life/3-D objects?

What is the computational approach? (Marr)

multi-stage approach that applies precise mathematical formulas

Primal sketch

image analyzed to determine "primitives": contours, object edges...based on natural constraints

2 ½ -D sketch

primitives are grouped (e.g., via Gestalt principles), and processed using depth information, depends on POV

3-D Model Rep

3-D component parts and their relations are determined, and matched to stored knowledge to identify object, independent of POV

which stage of computational approach is avail to conscious?

3-D model

Pros of comp approach

highly accurate, takes world knowledge into account

Con of comp approach

difficult to identify neural circuits for this processing

basis of rec by comp (biederman)

visual scene can be decomposed into basic elements that are constant

Principle of Componential Recovery

if an object's geons can be determined, then the object can be identified--even if the object is partially obscured

pros of rec by comp

PRO has well-defined features

PRO can handle variation, novel stimuli

PRO accounts for viewpoint invariance

cons of rec by comp

CON cannot handle context

CON may be too --objects also differ in their details (i.e. telling apart two different similar birds)

CON not all 3-D - objects easily decompose into parts (e.g., puddle)

What did Thouless find?

observers never produced correct projective view; there was always some shape constancy at work

* critical factor in shape constancy: knowing your viewing angle/orientation of the object

What did Biderman find with location?

observers did worse identifying familiar items in an impossible location (position violation) than in an unlikely location (probability violation) or normal location (no violation)

Perceptual set

predisposition to interpret a stimulus in a particular way, based on past experience or knowledge

Stroop effect

you can't help but read the words

Perceptual priming

perceiving and identifying an object is affected by previous experience with that object (or another one similar in form)

Schemas

mental structures that help us organize the world into a coherent, meaningful whole....LEARNED

Pareidolia

illusion or misperception involving a vague or obscure stimulus being perceived as something clear and distinct

parapsychology

scientific study of phenomena produced by living beings that cannot be explained by known scientific laws and forces

Ecological approach assumptions

emphasized evolution; all info required for perception is available in the environment (bottom-up);

the visual scene is analyzed in terms of whole objects & surfaces, not edges, colours, etc. that comprise objects; perception is direct, immediate "resonate to the stimulus"

behaviourist

Ecological approach is a _____ approach to perception

CHANGE

in optic flow, what rate is constant?

affordances

the possibilities for action offered by objects and situations

pro and con of ecological theory

identifying invariants adds to understanding;

perceptual processing exists; is important

lateral IP

represents location you plan to move your eyes to ("saccade planning area")

medial IP

represents area one can reach to; responds to visual and somatosensory input ("parietal reach region")

ventral IP

responds to moving nearby visual input/tactile input to face; important for feeding behaviour

anterior IP

represents important shape information for grasping objects; located between sensory and motor areas

Types of AIP neurons

motor dominant (action, not SEEING), visual-dominant (action LIGHT), visual and motor (CUE not action, in light more than dark),

visuomotor grip cells

(seeing and grasping)

Anterior IP into brain process...

AIP --> intraparietal sulcus neurons --> premotor and motor cortex

Mirror neurons (MNs): activated when

goal-directed action is performed, and when a similar action is observed

activation found in areas of human premotor cortex when

reading a literal phrase or watching a video of fruit being bitten or a pen being grasped

* but not by a metaphorical phrase

hypergoal

MNs require comprehension of the...

anterior intraparietal sulcus

less activity in ____ when viewing endoskeleton

Theory of mind:

cognitive capacity to understand that others have beliefs, desires and intentions different from one's own; and the ability to form a theory as to these mental states

Sally-anne test

how was the theory of mind tested?

anosognosia

denying of paralysis and that of other patients...implies mirror neuron system

Patient M.P.

Damage to temporal lobe, impairing ability to name objects, Identified object more rapidly when given a cue referring to the object's function

Information Provided by Motion

1. attracts our attention
2. provides information about the object's three-dimensional shape
3. provides information that helps us segregate figure from ground
4. provides information that enables us to interact with the environment

What handles motion?

MAGNO, not parvo

Reichardt Motion-Detector Circuit depends a lot on...

timing!

problems with Reichardt Motion-Detector Circuit

following a moving object with your eyes (pursuit): object is perceived as moving, even though retinal position doesn't change

sweeping your eyes across a stationary scene (saccade)--it appears to remain motionless

Solution to Reichardt Motion-Detector Circuit problem

take motion of eyes into account

\

Corollary Discharge Theory Based on:

image displacement signal, motor signal, corollary discharge signal, comparator

IDS

detects motion across retina

Motion will not be detected when...

there is both a CDS and IDS

Motion perception with Corollary Discharge Theory (4 ways)

1. observing afterimage in a dark room (retinal image stationary, eyes moving)
2. pushing on eyeball but keeping eyes stationary (muscles push back, causing CDS)
3. Pursuit: moving object, eyes move with it (e.g., tracking a car driving past: no IDS)
4. eyes paralyzed, try to move your eyes (retinal image stationary, but CDS sent) Eye cannot move but since the CDS is sent, you get a sense of the room moving

timing

apparent motion perception depends on

simultaneity

< 30 ms:

60 ms

continuous movement

\>200 ms

succession (completely different sets)

SOA Law:

changes in ISI and frame duration do not affect motion perception unless there is also a change in SOA

Motion correspondence problem

which elements in frame 1 correspond with which elements in frame 2?

Nearest-Neighbour Principle

select alternative that minimizes overall "traveling distance" of elements

Striate cortex and motion

direction of motion across small receptive fields

Middle temporal area and motion

direction and speed of motion

Medial superior temporal area and motion

locating and reaching for moving objects

superior temporal sulcus and motion

motion related to animals and people

spoke matched to another at rate of 90 °/frame correctly appears to rotate

clockwise

spoke matched to another at rate of 270 °/frame incorrectly appears to rotate

counterclockwise

identifying humans with point-light walkers is associated with increased activity in

superior temporal sulcus

pros of uncanny valley

✓ may provide framework for making robots/CGI characters more humanlike

✓ may contribute to an understanding of social interaction

cons of uncanny valley

❌shinwakan may be a confounding of strange/familiar and eerie/not eerie (MacDorman, 2006)

❌ criticized as pseudoscientific

❌ evidence equivocal

2 motion requirements (for film)

on-screen image must be flicker-free

slightly different frames must be shown in succession

short-range motion detection

low-level; occurs within small visual angles (15-20'); governs real motion of stimuli

long-range motion detection

higher-level; occurs in larger visual angles (20'-6°); mechanism for apparent motion

Does colour exist in the world?

No; only in your mind

crypsis

camouflage, concealment, or disruption with respect to the surrounding environment

concealing colouration

type of crypsis; animal has same colour as its environment

disruptive colouration

type of crypsis: pattern breaks up outline so one individual doesn't stand out; may also produce "motion dazzle"

mimesis

camouflage or concealment by imitation of another object

disguise

type of mimesis; animal looks like another, unimportant object to predator or prey

mimicry

type of mimesis; animal looks like other distasteful or dangerous animal

only ______ colours can be hues

spectral (colours in the rainbow)