PSYC 213 Chapter Notes - Chapter 3: Optical Flow, European Credit Transfer And Accumulation System, Parahippocampal Gyrus

108 views9 pages
Chapter 3: Perception
!
Introduction
Visual agnosia: unable to identify objects visually even though able to identify using other
senses (i.e. touch)
Perception: processing of sensory information such that produces conscious experiences and
guides action in world
Physiology of Visual Perception
Visual perception involves processing of info conveyed by light energy
Cornea: outer transparent tissue of the eye
Pupil: small opening through which light enters --size controlled by iris, which gives eye its
color
Retina: sheet at the back of the eye on which light is focused
o Contains photoreceptors, packed in fovea
Primary visual cortex: region responsible for early processing of visual signal
o Retinotopic: information falling on adjacent regions of retina processed in adjacent
areas of V1
Initially thought that this was only region responsible for visual info processing
o Later shown that interacts with other areas of the brain
Highly modular --some areas for motion, some color, etc.
Achromatopsia: selective damage resulting in color perception deficit
Akinetopsia: inability to perceive motion (motion blindness)
Visual Pathways
Ventral pathway: what pathway, inferior temporal lobe
o Processing characteristics of objects and assigning meaning
o Visual agnosia
Dorsal pathway: where pathway, parietal lobe
o Processing spatial aspects & motion/location
o Akinetopsia
Connections between brain areas are bi-directional
Feedforward sweep: propagation info V1 --> dorsal & ventral pathways
o Bottom-up influence
Re-entrant feedback connections: connections allowing propagation of information from
endpoints of ventral & dorsal pathways back to V1
o Expectations influence early processing, biasing perception
o Top-down influence
Perception relies on both bottom-up and top-down influences
Ventral Pathway & Object Recognition
V1 processes basic features of objects; cells in ventral stream process complex aspects
o Inferior temporal cortex (IT) identify entire objects
i.e. Fusiform face area responds selectively to faces
Prosopagnosia is deficit in recognizing faces, due to damage to FFA
Unable to combine features into complete percept
Unlock document

This preview shows pages 1-3 of the document.
Unlock all 9 pages and 3 million more documents.

Already have an account? Log in
o Temporal lobe
Parahippocampal place area for recognizing places
Extrastriate body area for body parts other than faces
Object Recognition
Pattern recognition: ability to recognize an event as an instance of a particular category of
events
o From computer science; computer able to identify configurations
o i.e. recognize words as meaningful and not just scribbles, recognize object with handle
as coffee mug
Recognizing configuration requires interaction between perception and memory
o Memory trace: trace that experience leaves behind in memory
o Hoffding function: process by which perception contacts memory trace
Feature Detection
Feature detection theory: detecting patterns based on their features / properties
o Selfridge
o Objects and events made of features that are used to identify them
o A lot simplified, but influential
Pandemonium is simple version of this model
o 3 levels:
o Bottom = image or cluster of data in which pattern of features presented
Features: size, color, shape, etc.
o Next level is cognitive demons who examine pattern
Each demon looking to detect particular pattern (i.e. one for apples, one for
baseball, etc.)
When demon sees pattern similar to one looking for, it shouts; the more similar, the
louder it shouts
o Top is decision demon who selects cognitive demon shouting the loudest
Choice determines what pattern recognizes
Pelli, Farell & Moore: looked at effect of contrast between letters in word and background
o Black letters on white background have more contrast energy
Experiment where vary word length (2-16 letters) & contrast energy
o Shown word for 200ms, then must identify which one was from list of 26 words of same
length
The longer the word, the higher the contrast energy required to identify word
o Letters (unlike overall patterns) are crucial features of visual system
o Too difficult to identify large number of letters in short time when signal is low
Squelching: tendency of NS to inhibit processing of unclear features
o Visual system focuses on reliability --doesn’t want to "guess" when not sure
Recognition by Components
Explains how feature analysis applies to real world objects (how breakdown 3d images into
simple components)
Recognition by components (RBC) theory: assume that image on retina is segmented --breaks
it down at points of concavity where curvature is discontinuous (corner) into basic components
called geons
o Biederman
Unlock document

This preview shows pages 1-3 of the document.
Unlock all 9 pages and 3 million more documents.

Already have an account? Log in
o Then geons compared with geon configurations stored in memory, and when match,
recognize the object
Experiment where vary number of geons used to depict given object & varied complexity of
object
o Show object quickly and then ask to name
o More geons = better recognition (more details)
80% for 2-3 geons; quite good
More complex obtains recognized more efficiently --counterintuitive since contain
more information so should take more time to process
o Show that breaking down objects into geons is crucial for object recognition
Template Matching
Template: guide used to bring piece of work into desired shape
Store templates in memory corresponding to standard forms of configurations seen
o Compare current configuration with representative/prototypical forms that have in
memory
i.e. Can recognize letter written with different fonts
o Involves comparing each a with the prototypical a that have in memory
o If match good enough, recognize the letter
Difficult to explain specific characteristics of model --template matches with identical and
similar enough patterns (not fixed)
o Criticized
o But good idea in that recognize thing because are similar to underlying prototype that
have
Hintzman: multiple-trace memory model
o Record traces of each individual experience in memory --No matter how often particular
event experienced, memory trace is recorded every time
Distinguishes between primary & secondary memory
o Achieve secondary memory using a probe
Snapshot of info / active representation of experience in primary memory that
activate memory traces in secondary memory
Probe goes from primary --> secondary memory, activating memory traces similar
to the probe (current experience)
o Activated memory traces return as echo to primary memory
Echo made from all activated memory traces, like choir
Get general impression, lose individual memory trace's properties
Theory used to explain results of study by Posner
Posner: participants shown distortions of prototypical patterns ("concept")
o Dots randomly moving away from position in prototype
o Do not see prototype themselves
Later shown prototypes, original distortions and some new distortions of prototypes
o Asked to classify these into various concepts
Results: participants good at classifying prototypical patterns even though had not seen them
before
o Sometime misidentified prototypes as patterns previously seen, even though only had
seen distortions of it
Unlock document

This preview shows pages 1-3 of the document.
Unlock all 9 pages and 3 million more documents.

Already have an account? Log in

Get access

Grade+20% off
$8 USD/m$10 USD/m
Billed $96 USD annually
Grade+
Homework Help
Study Guides
Textbook Solutions
Class Notes
Textbook Notes
Booster Class
40 Verified Answers
Class+
$8 USD/m
Billed $96 USD annually
Class+
Homework Help
Study Guides
Textbook Solutions
Class Notes
Textbook Notes
Booster Class
30 Verified Answers

Related Documents