Information for Human Perception

Image generates electrical signals from retina (bottom) visual receiving area of
the brain (up) [Bottom-up processing (information stimulating the receptors)]

Processing the environment from a perceptual system (top)  Enables people to

rapidly identify objects and scenes [Top-down processing (information based on
knowledge)]
 Perceiving Objects

 Hearing Words in a Sentence

 The ability to tell when one word in a conversation ends and the next
one begins is a phenomenon called speech segmentation
 The continuous sound signals enter the ear and triggers signals
that are sent toward the speech areas of the brain (bottom-up
processing)  If a listener understands the language, their
knowledge of the language creates the perception of individual
words (top-down processing)

 The likelihood that one sound will follow another within a word is
called transitional probabilities (e.g. pretty baby, less likely “ty” and
“ba” will be the same word) and the corresponding process is called
statistical learning

Conceptions of Object Perception

[1] Helmholtz’s Theory of Unconscious Interference
 The image on the retina is ambiguous (rapidly and unconsciously)

 People perceive the object that is more likely to have caused the
pattern of stimuli we have received (likelihood principle) and
corresponding process is called unconscious interference

[2] The Gestalt Principles of Organization

The whole is different than the sum of its parts (basic principles of Gestalt
psychology)

 This phenomenon is called apparent movement, and its

required conditions are:
 One light flashes on and off
 There is a period of darkness, lasting a fraction of a
second
 The second light flashes on and off at a different
position

 Good Continuation
 Based on the principle of 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
in such a way as to follow the smoothest path; AND
Objects that are overlapped by other objects are perceived as
continuing behind the overlapping object
 Pragnanz
  Based on the principle of good figure / simplicity:
 Each stimulus pattern is seen in such a way that the resulting
structure is as simple as possible

 Similarity
 Based on the principle of similarity (size, shape, colors or orientation):
 Similar things appear to be grouped together
 Although a person’s experience can influence perception, the role
of experience is minor compared to the perceptual principles
(different from Helmholtz’s likelihood principle which proposes
that our knowledge of the environment enables us to determine
what is most likely to have created the pattern on retina)
 Proximity (physical distance)

 Common Region

 Meaningfulness / familiarity
[3] Taking Regularities of the Environment into Account
 Physical Regularities
 Oblique effect (a phenomenon in visual perception where people have
better performance in detecting and discriminating horizontal and
vertical lines compared to oblique /diagonal lines)
 Light-from-above assumption (people usually assume that light is
coming from above, because light in our environment, including the
sun usually coming from above)

 Semantic Regularities
 The characteristics associated with the functions carried out in different
types of scenes (scene schema)

[4] Bayesian Interference (prior experience + perception  conclusion)

 Proposed that our estimate of the probability of an outcome is determined
by two factors: (e.g. cough / lung disease is most likely due to a cold ???)
 The prior probability, which is our initial belief about the probability
of an outcome (e.g. belief in having a cold, lung disease)
 The extent to which the available evidence is consistent with the
outcome (e.g. likelihood for cold / lung disease to cause coughing)

∴ The projection problem on the retina have a huge number of possible

objects and the most perceptual situation with prior probabilities based on
 our past experience

[5] Comparison of 4 approaches

 Helmholtz, regularities, and Bayes are based on our past experiences in
perceiving  Top-down processing
Gestalt acknowledges that perception is affected by experience but argued
that built-in principles can override experience  Bottom-up processing

 In modern psychology, people point out that the laws of organization could
be created by experience (e.g. the principle of good continuation has been
determined by experience with the environment)

[6] Theories for object recognition

 Template matching (to match the perceived image to mental images stored
in long-term memory) [recognition depends on finding a “good” match]
 Template matching with transformation (rotation + moving +
changing size)

[7] Structural-description theories

 To present an object by describing the structures and the relations among
them [recognition depends on activating the correct set of structures and
relations  matching the input list of structural description in memory]
 Recognition by components (RBC)
 Stored as geons (e.g. cylinders, cones, rectangular blocks …)
 Objects are represented as a list of “geons and their relations”
(e.g. Ice-cream = sphere on top of cone)
[8] Face Perception
 A cortical region for face process (FFA)
 Involves holistic 全面的 processing and inversion effect

 Well-trained objects also activate FFA

Neurons and Knowledge about the Environment

[1] Neurons that Respond to Horizontals and Verticals (Evolutionary perspective)

 People are more sensitive to horizontals / verticals than other orientations
(physical angularities: oblique effect) because more neurons respond to that,
and this is called theory of neuron selection:
 Characteristics that enhance an animal’s ability to survive, and
therefore reproduce, will be passed on to future generations

[2] Experience-Dependent Plasticity (learning perspective)

 Demonstrated by brain imaging technique of fMRI  finding that there is
an area in the temporal lobe called fusiform face area (FFA) that contains
many neurons that respond best to faces:
 After training, FFA contains neurons that respond not just to faces but
to other complex objects as well

 Experience-Dependent Plasticity can override neurons that respond to

 vertical / Horizontal (e.g. rearing kitten experiment)

Perception and Action: Behavior

[1] Movement Facilitate Perception

 Seeing an object from different viewpoints provides added information that

results in more accurate perception

[2] The interaction of Perception and Action

 Perception and Action: Physiology

[1] “What” and “Where” Streams

 Brain ablation / lesioning
 An experiment to determine the function of a particular area of the
brain by removing part of the brain (using monkey because of the
similarity of the monkey’s visual system to that of humans and monkey
can trained to demonstrate perceptual capacities):
 Object discrimination Problem (demonstrate “what” pathway)
 Landmark discrimination problem (demonstrate “where”
pathway)

 “what” pathway / ventral1 pathway

 “where” pathway / dorsal2 pathway

1
Refers to the opposite of dorsal
2
Refers to the back surface of an organism
[2] Perception and Action Stream
 Through a neuropsychological approach (studying the behavior of people
with brain damage)
 Pathway from the visual cortex  temporal lobe (what / perception
pathway)
 Pathway from the visual cortex  partial lobe (where / action
pathway)

[3] Mirror Neurons

 Neurons responding to “observe experimenters picked up a piece of

food” = Neurons responding to “picking up the food”  mirror
neurons (** the type of object is not important. E.g. “others object
but not “food”)

∴ Mirror neurons may involved in determining the goal or intention

behind an action
 The Perceptual Process

[1] Human Perception

 Perception is experiences resulting from stimulation of the senses

 Sensation occurs when sensory receptors detect sensory stimuli and

transform the energy in the environment (e.g. light energy …) into nerve
impulses [objective]
 Perception is the process to interpret the sensations (e.g. electrical signals)
and form subjective conscious experiences
 Visual Perception

[1] Structure of eyes

1. Iris controls the size of the pupil
2. Lens focuses light into the retina
3. Ciliary muscles control the thickness of the lens to
focus at different distances
4. Vitreous body is a gel-like substance fills the eyeball
5. Retina is a structure at the back of eyeball containing
photoreceptors

[2] Stages of visual processing in the visual system (Stimulation from eyes  Brain)
1. Retina
 Images projected on the retina are inverted both vertically and
horizontally
 Light passes through all the neurons before reaching the
photoreceptors

 Structure:
 Fovea
 Central part of the retina (contain almost no ganglion axon
or blood vessel)

 Periphery
  Surrounding of retina

 Optic nerve
 Axon of neurons (carry information from retina to the brain)
 There is a point where the left and right optics nerve cross
(optic chiasm)

 Blind Spot
 Location of the optic nerve leaves the eye
 Does not contain photoreceptors
 Contains photoreceptors:
 Photoreceptors transduce light energy to neural signals:
 Outer segments of photoreceptors contain visual pigment
molecules (absorb lights of specific ranges of wavelengths);
Visual pigments have two components (opsin and retinal)

 Two different photoreceptors:

 Rods (vision for dim light)
 Cones (vision for colors)

2. Retinal ganglion cells (RGC) (still inside retina)

  Process of Visual Transduction which involves opsin3 and retinal4
 Retina changes its shape (isomerization)  electrochemical
changes on the photoreceptors
 Humans are able to detect a single-photon stimulus

3. Lateral geniculate nucleus (LGN)

 A small brain structure in the thalamus (part of midbrain) that relays
visual signals from the eye to the brain
4. Cerebral cortex
 Primary visual cortex / V1 / striate cortex / Brodmann area (BA) 17
 Located at the back of the cerebral cortex (occipital lobe)
 First visual-processing area in the cerebral cortex

 Contain neurons that are highly-selective to specific kinds of

primitive visual features


[3] Dark Adaptation (~ 30 mins)

 The process of increasing sensitivity after staying in the dark for a certain
period of time

 Based on Duplicity Theory5 (such that rod and cone mechanism will not
have an abrupt transition)

3
A large protein
4
A light sensitive molecule
5
States that above a certain luminance level (~ 0.03 cd/m2), the cone
mechanism is involved in mediating vision: photopic vision; Below this level, rod
mechanism comes into play by providing scotopic (night) vision
  For the first 10 mins, you are able to differentiate colors (dominate by
cones) and after that you observe all colors as grey (Purkinje effect:
dominated by rods)

Selectively and Adaptation

[1] Selectively of V1 neurons

 V1 neurons are highly selective in responding to specific stimulus features
(e.g. vertical …) and we call these features as orientation-selective

[2] Selective adaptation

 If a neuron is exposed to the stimulus it prefers for a long time, its firing
rate decreases in general (adaptation: “the neuron fatigue” / neural firing
reduces most at the preferred stimulus)
 Change in neural activity affects our perception  adaptation effect
can be measured psychophysically

 Adaptation effects depend on the specific stimulus and the tuning

preference of the neuron (selective)
 Example: Negative Afterimage / “Waterfall” effect / selective rearing of
kittens
 Depth Perception

[1] Oculomotor (information from muscles in the eyes)

 Accommodation (shape of lens changes when we focus on objects at
different distances)
 Vergence (inward / convergence and outward / divergence movement of the
eyes when er focus on objects in different depths)

[2] Monocular (information obtainable from the retinal image of one single eye
only)
 Pictorial cues (no motion)
 Occlusion
 A non-transparent object blocks the view of part of another object

 Relative height
 Object below horizon (higher position in retinal image)  Further
away
 Object above horizon (higher position in retinal image)  Further
closer

 Relative size
 All things being equal, a closer object subtends a larger visual
angle on the retina than a further object does:
 Larger objects in a picture appear closer
 Smaller objects in a picture appear further away
 Familiar size

 Linear perspective (perspective convergence)

 There is a vanishing point at which parallel lines receding in
depth converge

 Texture gradient
 A depth cue based on the geometric fact that items of the same
size form smaller images when they are further away
(combination of relative size and relative height)
  Aerial perspective (atmospheric perspective)
 Based on the implicit knowledge that light is scattered by the
atmosphere
 More light is scattered when we look through more
atmosphere (∴ more distant objects appear fainter, bluer, and
less distinct)

 Shadows

 Kinematic cues (required motion)

 Motion parallax
 When the observer moves, objects closer to the observer move
faster compared to images further away
 speed-difference information to calculate the distances of
 objects in a moving scene

 Optic flow
 Accretion and deletion

[3] Binocular (information from comparing the retinal images on both eyes)

Real life examples

[1] Blind Spot

[2]
Use single eye to focus “F” and you will find that
the red spot disappears as yellow background
instead of a black dot (perception fill in)

Motion-induced blindness
 A visual phenomenon in which stationary objects in a person's field of
vision seem to disappear or fade out when surrounded by a moving pattern.
 Explanation: When the brain is focused on tracking movement, it can
sometimes overlook or suppress the perception of stationary objects.
This is thought to be a result of neural mechanisms that prioritize
dynamic, changing stimuli over static ones, which can be less critical
for survival.

[3] Figure-ground segregation

 Segregate the object (figure) from the background (ground)
 Size (figure always smaller than background)
  Symmetry

[4] Physical regularities

 Light from above assumption

[5] Constancy
 Perceptual constancy
 The ability to recognize the same stimulus across different conditions
(e.g. lightness, color, size …)

Color constancy
 Size Constancy

 The phenomenon that our perceived sizes of objects remain

constant even when their retinal images change due to changes in
depth (closely related to our knowledge, expectation …)

[6] Müller-Lyer illusion

 Explanation:
 Conflicting Cues Theory
 The illusion arises from the conflict between the actual length of
the lines and the length implied by the contextual cues of the
 arrowheads. The brain resolves this conflict by averaging the two
lengths, leading to the perceived difference.

 Gestalt Principles (principle of closure)

 Our brain tends to perceive complete shapes. The arrowheads
might create the perception of a larger or smaller enclosed space,
influencing our judgment of the line length.

[7] Ponzo illusion

 Explanation:
 Perspective and Depth Cues
 The converging lines in the Ponzo illusion resemble railroad
tracks or roads that converge in the distance. Our brain interprets
these lines as a depth cue, indicating that the top part of the image
is farther away than the bottom part.
 Size-Distance Scaling
 When our brain perceives an object as being farther away, it
automatically compensates for this distance by scaling up the
perceived size of the object. This is known as size-distance
scaling. In the Ponzo illusion, the top horizontal line is perceived
as being farther away, so our brain scales it up, making it appear
longer than the bottom line.

[8] The Ames Room

 Features:
 Irregular shape
 The walls are trapezoidal, with one corner being much farther
away from the viewer than the other corner. The ceiling and floor
are also sloped.

 Forced Perspective
 When viewed from a specific vantage point (usually through a
peephole or a small window), the room appears to be a normal
rectangular space. This forces the brain to interpret the scene
using familiar cues of perspective and size constancy.

 Explanation:
 Visual cues
 The illusion relies on the brain's reliance on visual cues and
perspective to judge size and distance. The irregular construction
of the room manipulates these cues, causing the brain to perceive
the room as normal and the objects within it as changing in size.

 Contextual Information
 The brain uses contextual information from the room's design to
maintain a sense of normalcy. This leads to the misperception of
size, as the brain attempts to make sense of the conflicting
information.