Topic 3 - Perception Flashcards
There are a myriad of problems associated with perception, mainly due to the
sensations that we receive being incomplete or ambiguous.
Is perception active or passive?
Active
What are the 2 problems with perception?
1- inverse projection problem
2- size
inverse projection problem identifies that for any
2D retinal image there are a potentially infinite number of 3D shapes that could have produced that activity
Starting point for visual perception is the
registry for sensory info at the back of the retina.
Visual perception is the interpretation of
retinal info
Size effecting perception
2 hypothesis
1) tiny object in hand
2) object is far from distant
Example of size being an issue for perception?
Moon illusion
Moon illusion
Moon looks larger when there is no horizonal objects. When horizon is clear moon looks a lot larger.
Transitional probabilities
We have learnt that certain pairs of sound are more likely to belong to the same word (par-ty) while others more likely belong to different words (ty-go).
Which sounds predict others sounds? And what are there probabilities?
Bi —-> Da —-> Ku = 1.0
Pa —-> Do —-> Ti = 1.0
Go —-> La —-> Bu = 1.0
Ku —> Pa, Go, Bi = 0.33
8 month old infant are more interested in
Parts of words rather then whole words. More interested in the starts and ends of words.
Perception is entirely
Continuous
The more predictable the words are, then infants are more
Interested in them
What are the general principles that apply to all sensory systems?
1) transduction
2) organization
3) symbolism
4) selectivity
5) hierarchies
6) tailoring
7) thresholds
8) decision rules
Sensation is what
Perception is what
- Sensation is bottom-up processing
- Perception is top-down processing
Sensation is relieved from the outside world via… hows it processed?
Sensory systems and process through bottom-up processing
What is perception and what does it require?
It’s the best prediction of what’s happening.
It requires previous information and experience for top-down processing.
What does decision rules emphasize
The difference between sensation and perception
Transduction
The outside world (distal stimuli) sends us signals which our sensory transducers (eyes, ear, nose, tongue, skin) encode and covert into bodily sensations (proximal stimuli).
The process of transduction may take place over
vast distances (far senses) or has to take place at the point of contact (near senses).
Symbolism
Characteristics of the object or sensation are recoded neurally and are represented symbolically. Characteristic like smell or colors.
For symbolism, certain brain states give rise to
Specific perceptual experiences
Organization
Neurons at each stage of processing are arranged in a highly organised way, according to the nature of the sense.
- auditory organization
- visual organization
Auditory organization
Frequency selective in the basilar membrane
- low frequencies at apex (20Hz)
- high frequencies at the base (20KHz)
Visual organization
Spatial location selectivity in the striate cortex
- fine grained analysis at focal point,
- cruder analysis around periphery
Selectivity
There is selectivity both at a sensory level in terms of transduction specialisation, and also at a neural level where individual cells prefer certain types of stimulation above others.
TOUCH -
AUDITION -
VISION -
TOUCH - Pacinian corpuscle (squashing sensation)
AUDITION - Inner hair cell
VISION - Orientation selectivity
Which eye parts are involved in selectivity and what do they each do?
- Sclera = tough white bit for protection
- Cornea = transparent area that allows light in
- Iris = controls size of pupil and gives eye colour
- Pupil = Opening at the centre of the iris
- Lens = Transparent structure that bends to focus light
- Retina = Light-sensitive cells at the back of the eye
Selectivity in colour perception
- 3 kinds of cones, B, G, R.
- mixing these cones give different colours like purple or orange.
- fewer B cones then R and .
Dichromates only have
2 cones, most common is lack of G or R.
Terachromat
Four cones (goldfish), UV cones
HIERARCHIES
Perceptual processing involves a series of stages. Each stage one representation is transformed into another. Each sense contains multiple, hierarchically organised processing stages. Sensory systems also have ways for higher levels to feedback to earlier levels.
TAILORING
HAVING SENSORY TRANSDUCERS THAT ARE SPECIALLY TAILORED TO PERFORM A SPECIFIC TASK IN A SPECIFIC ENVIRONMENT.
THRESHOLDS
- absolute threshold
- just noticeable difference (JND)
- sensory adaption
(absolute threshold)
A minimum amount of stimulation is required to evoke a perceptual sensation
(just noticeable difference; JND).
A minimum amount of differential stimulation is required to note the change between sensations
(sensory adaptation).
The same level of stimulus intensity does not generate the same continuous level of response
DECISION RULES
Frequently, perceptual information is incomplete or ambiguous. Therefore, stimulus processing from the bottom-up cannot account fully for perception. Top-down decisions also need to be made.
(Rotating snakes illusion picture)
CLUES TO RECOGNITION
DECISION RULES
Sensory information from the outside world does not have a single interpretation- we live in an inherently ambiguous environment. Retinal (or any other kind) of ambiguity is resolved via a number of individual and shared expectations.
Depiction making involves
- unconscious inferences
- generic viewpoint
- unconscious inferences
In addition to Prior experiences , we make shared unconscious inferences to interpretate sensory info
Because of ________ we are more likely to choose some interpretations then others. Why?
Generic Viewpoint
Because we naturally distrust unique perspective and favour more generic ones.
Trichromacy theory
establishes the different cones available to us.
Opponency theory:
in color perception, when certain colors are perceived in certain locations, not only are there predictions made about what colors that is, but there’s the suppression of info on what color that CANT be there.
- (You don’t describe it as reddy-green or bluey-yellow. You say bluey-green, or yellowey-red)
- Why is this? Because of the way our cone pigments are set up and communicate, you can never have green and red in the same location, in the same way you cannot have blue and yellow in the same location.
Which colors are opponent to one another?
Red opposes green, and blue opposes yellow.
Heavy does of red results in ?
Heavy dose of yellow results in?
Heavy does of red results suppression of the likelihood that green will be perceived at the same spot and time.
Heavy dose of yellow results in the suppression of the likelihood that blue will be perceived at the same spot and time.
Going from a heavy does of one color results in what
Results in the phone t color appearing in a pale shade.
Bright red = light green
Bright blue = light yellow
Bright green = orange
Bright yellow = light blue
GESTALT PRINCIPLES
Gestalt Psychology (30 years on from Helmholtz) proposed a set of perceptual organization principles, specifying the shared ways in which many of us resolve sensation. These are heuristics not algorithms, so might not be right every time.
What are the gestalt principles?
- pregnanz
- similarity
- proximity
- closure
- good continuation
- common fate
- familiarity
- light from above
- occlusion
Can these principles combine? What are the combos if so?
YES
- (good continuation + common fate)
- ( closure + good continuation)
- (closure V proximity) these compete
Is retinal info 3D or 2D ?
2D
SIMILARITY
Similar things appear to be grouped together
PROXIMITY
Things that are near to each other appear to be grouped together
CLOSURE
The propensity to ‘see’ closed rather than open forms
- using negative space to imagine lines that aren’t there to connect the object into itself.
Like the WWF panda
GOOD CONTINUATION
Points that, when connected, result in straight or smoothly curving lines are seen as belonging together, and the lines tend to be seen as following the smoothest path
- when lines connect to a certain point, the image/ object likes proper. But going to far distort the object to looking fucked up.
COMMON FATE
Things that are moving the same direction appear to be grouped together
FAMILIARITY
Grouping is more likely when we recognise what we see
- once its recognized you cannot unsee it
LIGHT FROM ABOVE
Most light in our environment comes from above
- Turning a hole into a bump when image inverts.
OCCLUSION
When a larger object is partially covered by a smaller occluding object, we see the larger object as continuing behind the smaller occluder
- instead of seeing 2 horses, you see one really stretched out horse.
Combining two images together with the lines lining up and both sides move at the same time is the combination of which gastlt principles?
Good continuation and common fate.
Closure + good continuation
When a shape has a portion missing but you can still tell what the shape is.
Closure OR proximity
These principles compete with one another to each produce a different kind of perception.
)(. )(. )(. )(. )(. )(
Two competing interpretations, the interpretation that you see depends on which principle you choose to use.
- if you use proximity, you group the opposite brackets together = at least 5 hour-glass shaped objects.
- If you use closure, you group closing brackets together = 4 circle shaped objects.
Which gastalt principles compete
Closure and proximity
Moving from one perception to another depending on which gastalt principle you choose indicates
Perceptual decision making (top-down processing)
Bottom-up processes can try to
supply us with ‘untainted’ information about our sensory world.
Top-down processes can help or hinder the
interpretation of our sensory world via expectations.
Bayesian Inference restates the
Helmholtz idea and quantifies how we might make a decision about (perceptual) ambiguity. This is based on the statistical and semantic regularities we have acquired about our world.
Bayesian Inference formula
Prior X likelihood = probable conclusion
- prior = belief about the frequencies
- likelihood = causal consequences and their probabilities
Patient SM suffered visual agnosia
the inability to recognise objects or faces. SM was trained with a novel collection of objects known as Greebles (on the left).
After a prolonged period of training, SM was able to correctly recognise different ‘sexes’ and ‘families’ of Greebles. The fusiform face area became more excited about Greebles, but not to faces.
the ________ is thought to be flexible in accommodating expertise regarding visual objects.
Fusiform face area
individual environment exposure, our increased prior to vertical and horizontal (but not diagonal) lines might be related to
natural selection and our ancestral exposure to Vs and Hs.
Visual information originating at the Occipital Lobe might take one or two independent routes:
- DORSAL PATHWAY To —> Parietal Lobe For object location (where)
- VENTRAL PATHWAY To —> Temporal Lobe For object identification (what)
If the Parietal Lobe is ablated (removed) then
landmark discrimination becomes difficult due to disrupted where circuitry
If the Te m p o r a l L o b e is ablated (removed) then
object discrimination becomes difficult due to disrupted what circuitry
Patient DF had Te m p o r a l L o b e damage due to carbon monoxide poisoning
So this should not affect
action- based responses, but should affect non-action-based responses.
PRAGNANZ
Every stimulus pattern is seen in such a way that the resulting structure is as simple as possible
Ie. the most simplest solution is chosen.
