Perception Flashcards
Defining Perception?
Possible definitions:
– the act of becoming aware of something through our senses
– Pattern recognition: ability to detect meaningful patterns in the environment (faces, objects, etc.)
Why study perception?
Perception is not always equivalent to reality
Sometimes perceptual errors offer a clue to how visual cognitive processes work (such as optical illusions)
Understanding perception helps designers create user-friendly displays
Inverse Projection Problem?
Determining the object responsible for a particular image on the retina
Objects can be hidden or blurred
How do we got rom this 2 dimensional picture to the experience of 3 dimensions?
View Invariance?
If the primary visual cortex only gets a 2D map from one
angle, how do we know what an object looks like from other
angles?
How do we know what the bottom of a chair looks like from seeing it from another angle?
Size Constancy?
How do we know something is the same size at different
distances, since when it’s further away it is smaller on our retina?
Perceptual processes?
Bottom-up processing:
Basic elements come together until the mind reaches a higher level of understanding
Top-down processing
Prior knowledge or context are used to analyze incoming information to inform perception
Helmholtz’s Theory Of Unconscious Inference? and what principle?
Some of our perceptions are the result of unconscious assumptions we make about the environment
Likelihood principle-we perceive the world in the way that is “most likely”based on our past experiences (Top down inference using prior knowledge)
Gestalt Principles of Perceptual Organization - Similarity?
Elements that look similar will be perceived as part of the same form or group
Gestalt Principles of Perceptual Organization - Proximity?
Elements that are close together will be perceived as a coherent group
Gestalt Principles of Perceptual Organization - Closure?
Humans tend to enclose spaces by completing a contour and ignoring gaps in a picture
Gestalt Principles of Perceptual Organization - Common Fate?
If two or more objects are moving in the same direction and at the same speed, they will tend to be perceived as a group
Gestalt Principles of Perceptual Organization - Symmetry?
Images that are perceived as symmetrical are experienced as belonging together
Gestalt Principles of Perceptual Organization - Good Continuation?
People tend to connect elements in a way that makes the elements seem continuous or flowing in a particular direction
Bayesian Inference?
Our estimate of the probability of an outcome depends on…
Top-down perception using:
1. Prior probability (“prior”) of the outcome
2. How much the available evidence is consistent with the outcome (“likelihood”)
From eye to cortex? (4)
Eye -> Optic nerve -> Lateral Geniculate Nucleus -> Visual Cortex
What and Where Pathways?
Ventral Stream: What (Object recognition)
Primary auditory cortex
Dorsal Stream: Where/How (Movement and spatial relations)
Primary visual cortex
Inferotemporal Cortex (IT)?
- Front end of ventral stream
- Cells selective for object categories
- Some regions selective for faces
- “View invariant”
Object identification: 2D image to 3D objects? (Ventral Stream)
Common to many theories:
- Reconstruct representations (or structural descriptions) of objects
- Match them to some kind of temple
Bottom-up Processing for object identification?
Extract basic features from a stimulus and compare to known patterns of stimulus features stored in long-term memory
There are at many different descriptions of how we employ basic bottom-up processes
- Distinctive Features Theory
- Recognition by Components Theory
- Template-Matching Theory
- Prototype Theory
Distinctive Features Theory?
All complex perceptual stimuli are composed of distinctive and separable attribute called features
- Allow observers to distinguish one object/person form another
Pattern recognition is accomplished by mentally assessing the presence or absence of a checklist of curial features
Recognition by Components Theory?
Describes the pattern recognition process in terms of how people recognize 3-D objects—by identifying the building blocks that make up the objects
Basic elements are composed of an alphabet of 36 primitive shapes, called geons
- Geometic icons
- Building blocks for identifying #D objects
(solves the view invariance problem)
Template Matching Theory?
- We store an unlimited number of patterns, literal copies corresponding to every object that we have experiences
- These patterns are labeled with the name of the object
- New instance matched to stored templates
- Informs us of the name
Problems with template matching theory?
- It is highly impractical when the set of possible patterns is very large
- Inefficient and fails to account for our ability to recognize new objects
- Example, How do we manage to read the handwriting of a person we just met?
Prototype Theory?
Template is not a literal match with an object, but is an average or typical instance of the many different views of that object
This average is called a prototype andthe theory that uses them for pattern recognition is called prototype theory
Pattern recognition occurs when the features of the object to be recognized overlap with the features of the prototype
Prototype Theory does not require?
An exact match between the object and the prototype
The storage of patterns for every possible view of an object
Prototype Theory test: For each prototype, researchers created individuals that varied in their degree of similarity to the prototype
Stage 1: learning phase
Stage 2: judgment phase (old/new face)
Result?
The prior faces that they saw they had higher confidence that they’d seen before. The faces they had not seen before had a low confidence of seeing them before.
But there was 100% confidence that they’ve seen the prototype face, even though they never seen it. Moreover, the more similar the faces were to the prototype the higher was the confidence that they’ve seen them before.
Facial Prototypes?
Faces are highly varied and, except in rare circumstances, no two are alike
Yet they share significant similarities
- Always contain the same set of parts
- Always appear in the same basic configuration
The Thatcher Illusion?
Typical face processing: gestalt-like manner, using spatial relations among the face parts to construct a holistic view
When the face is rotated, we switch to noticing the individual facial features—the component information (more common for non-faces)
Facial Prototypes -> We’re better at recognizing faces when they are right side up and looking forward then when upside down?
We have only developed a prototype for right-side-up faces
The fact that we are poor at recognizing upside-down faces suggests that we don’t have practice seeing faces in that position
Neural circuitry for face recognition? (FFA) which is inside the?
One of several areas of the brain that play a key role in face recognition is a region of the fusiform gyrus called the fusiform face area (FFA)
FFA is inside fusiform gyrusis inside inferotemporalcortex (Ventral stream)
FFA is part of face-selective network
Functional imaging in monkeys has revealed nine faceselective
patches in temporal and prefrontal cortex
Prosopagnosia?
Inability to recognize familiar faces (face blindness)
Not a result of visual, intellectual, or memory difficulties
May be able to recognize people by their voices, names, or by superficial features (e.g., hair color)
Rely on more featuralprocessing
Top-down processing?
Using knowledge stored in long-term memory to further analyze incoming information in order to complete the identification
- Expectations
- Context
- Interpretations
- Biases
What is a feedforward process?
V1 -> V2 -> V3 ->IT
When signal or information get passed forward from one area to the next
Bottom up processing involves mostly feed-forward processes as infraction gets passed on from V1 forward
V1 <- V2 <- IT <- PFC
When signal or information gets passed backward from a higher level (further from V1) area to the a lower level one
Feedback processes are important for top-down influences on face and object recognition
Key Brain Regions? V1 or primary visual
cortex, IT or inferotemporal cortex, & Fusiform Face Area (FFA)
V1 or primary visual
cortex, which is where
information from the
eyes reach the cortex
and the beginning of
both dorsal and
ventral visual streams
Prefrontal cortex or PFCis important for taking context into account in top-down influences on face and object processing
IT or inferotemporal cortex which is important for
recognition of whole objects
Fusiform Face Area (FFA): A region of the fusiform gyrus in
IT that is an important part of the face network
Top-down contributions study.
Used fMRI and MEG to
determine when and
where in the brain object
recognition occurred
Information got to the
frontal lobes for recognized objects before it reached
inferotemporal cortex (IT)
Conclusion?
The frontal lobes predictan object’s identity and help fine tune the representation in IT
Speech perception? (3+ principle)
Top-down processing influences our perception of language based on our individual experiences with manage
Some perceptions are the results of unconscious assumptions we make about the environment -> We use our knowledge to inform our perceptions
We infer much of what we know about the world
Likelihood principle:
We perceive the world in the way that is “most likely” based on our past experiences
Speech segmentation?
The ability to tell when one word ends and another begins
Transitional probabilities?
Knowing which sound will likely follow another in a word
Statistical Learning (Saffran)
Infants 8 months or younger can learn transitional probabilities
Invented nonsense words using a regular pattern and unregular patterns to form the word -> low and high transitional probabilities
8 month old infants showed more interest in novel words, with low transitional probabilities
Regularities in the Environment? & Oblique effect?
Common physical properties of the environment
Oblique effect: (means an angled line)
We perceive verticals and horizontals more easily than other orientations
Light-from-above-assumption?
We assume light comes from above because this is common in our environment
We perceive shadows as specific information about depth and distance
Semantic regularities and Scene schema?
- Semantic context that helps us perceive contents faster
- Top-town context provides perceptual processes with expectation what is coming
Optic Ataxia?
A clinical disorder that causes problem with “visually-guided” behaviour stemming from impaired representations in the dorsal (where/how) pathway. Demonstrating that it is a distant process from identifying what the object is (ventral pathway).
Apperceptive Agnosia?
Failure of perception in which some basic visual functions (acuity, color, motion) are preserved, while others, (shape, angle, size) may be disrupted
Damage to the ventral (what) pathway
Associative Agnosia?
Apperceptive agnostic people can have difficulty recognising objects, but it is due to deficits perceiving simple visual information (angle/width). They still have preserved object knowledge
-> Inability to recognise objects despite intact perception of basic visual characteristics, usually caused by damage to the ventral stream
Gist perception?
- Abstract meaning of a scene is automatically extracted
- Takes ~150 milliseconds
- Happens before conscious attention can be effectively deployed to explore the scene
Based on little information:
- Influenced by colour, orientation, crowding/openness
- Doesn’t necessarily involve object representations or details
Scene Layout Perception?
- Similar to gist perception, but takes a few seconds more
- After a few seconds, we can perceive the spatial layout of objects in a scene
- Repeated layouts are learned implicitly (without awareness)
Statistical Estimation/Ensemble Perception?
When there are many objects, the perceptual system can quickly extract “ensemble” properties
- Such as mean size, orientation, velocity, direction, facial expressions and more
Takes less than 50 milliseconds
Possibly related mechanism:we can rapidly estimate the percentage of the group that is a certain colour or orientation
Pattern Detection?
- Perceptual system can detect trends or outliers on a graph
- Related to ensemble perception
- Related to gestalt grouping by proximity
Grouped objects are easier to perceive -> reduce clutter
Special Perception for Visuomotor Control?
- There may be more accurate perception for motor-planning in the lower visual field
- Based on the idea that there are separate perceptual systems for conscious object identification and implicit control of action (e.g. reaching or pointing)
Association Formation/Operant Conditioning?
- Rensik mentions a situation in which displays “…help a user move a mouse…[without] awareness of control…” (p.33)
- Perhaps this “auto-pilot” can be accounted for by simple habit formation by repetition or operant conditioning
- Cross-platform design uniformity achieves this
