LECTURE 5 Flashcards
VISION
Construction of the world based on incoming information provided by light interacting with
photoreceptors on the retina.
* Most important/informative sense for humans
* Greatest amount of sensory cortex is devoted to vision
* Conscious experience dominated by visual information
EARLY NOTIFICATION VS RICH DETAIL
Tradeoff between accuracy and speed
LIGHT
Comprised of photons – particle and wave
* Color (and visibility) of light related to its
wavelength
* Brightness of light related to its amplitude
RETINA
Light stimulates photoreceptors at far end of retina
* Neural signals converge onto ganglion cells through bipolar cells
* Ganglion cell axons group into the optic nerve, which projects into brain
RODS
Low light vision (scotopic)
* Sensitive to fast motion
* Low resolution; poor edge detection (high convergence)
* Concentrated at peripheries of eye
* ~120 million per eye
CONES
Requires bright light (photopic)
* Three types – short, medium, and long – correspond to wavelength most sensitive to
* High resolution; good edge detection (low convergence)
* Concentrated in the fovea (middle of retina; focus point)
* ~6 million per eye
CONES VS RODS
Although the rods greatly outnumber the cones, each cone provides a much larger relative input to visual cortex due to neural convergence
* ~90% of brain’s input originates from the cones
DOES RIGHT EYE GO TO LEFT BRAIN AND VICE VERSA?
No. Left visual field projects to the right hemisphere of
visual cortex and right visual field projects to the left hemisphere of visual cortex
THROUGH THE BRAIN
Information from the optic nerve splits to both hemispheres in the optic chiasm
* Routed through the lateral geniculate nucleus (LGN) in the thalamus
* Projected to primary visual cortex, area V1 in the Occipital lobe
LGN NEURON (THALAMUS) COMPRISED OF
Parvo and Magno cells
PARVO CELLS
- Connections from foveal cones
- Color sensitive
- Detail and edge detection
- Good spatial resolution
MAGNO CELLS
- Connections from peripheral rods and cones
- Sensitive to motion over wide areas
- Bad spatial resolution
- “Fast track”
VISUAL CORTEX
- Area V1 (primary visual area)
- Back of occipital lobe
- Individual neurons responsive to edges of specific orientations
- Retinotopic organization
- Cortical magnification
PARIETAL LOBE PATHWAY
Dorsal pathway (spatial location and action)
TEMPORAL LOBE PATHWAY
Ventral pathway (characteristics of objects)
COLOR - PHYSICAL AND PERCEPTION DEFINITION
- Physics definition
- Wavelength of light
- Can be emitted from a light source or reflected off a surface
- Perceptual definition
- Experience of “color”
- Based on wavelength – but only partially
- Perceptual experience of color changes based on
brightness and hue of light source and lightness of object
COLOR MIXING
Additive color mixing
* Mixing light
* Every new color “adds” emitted color
* Colors can be defined as a combination of red, green,
and blue (RGB)
* Photoshop RGB color picker
CONES AND COLOR
- Three types of cones, corresponds to size of
wavelength most sensitive to: Short (Blue), Medium (Green) & Long (Red) - Combination of information from three cones is
unique for all visible wavelengths - S-cones are less common, and absent in fovea
- M-cones are a more recent development (cf. dog
vision)
OPPONENT-PROCESS THEORY
Color “opposites”
* Red-Green, Blue-Yellow, Black-White
* Color adaptation effect
* Neurons “fatigue” when responding to same stimulus
* When stimulus is removed, “opposite” is perceived
COLOR WHEEL
Colors opposite from each other are opposing
Opposing colors “pop” when combined = Color contrast
BRIGHTNESS
Intensity of light; luminance
* Depends on lighting, transient
LIGHTNESS
Degree to which a surface reflects light;
reflectance
* Quality of an object itself, stable
LIGHTNESS CONSTANCY
Interplay of lighting and lightness determines brightness (lightness constancy)
* Both a white sheet of paper in a dark room (a) or a grey sheet of paper in a bright room (b) have the same brightness
* Therefore - a must be a lighter surface than b since brightness is the same, but a is viewed with less overall light
COLOR CONSTANCY
This same process occurs when evaluating the color of objects in different lighting contexts
* Colors reflect different spectrums under different colored lights
* Like lightness constancy, need to figure out the color of the object itself, so need to “subtract” ambient lighting
SIZE OF AN OBJECT
The size of an object on the retina is determined by
both the size of the object and the distance it is
from the observer.
* To account for this, objects can be measured in terms of visual angle.
SCENE
External world of objects (3d)
IMAGE
Light reflected by the scene and picked up
on retina (2d)
PROBLEM OF INVERSE OPTICS
Theoretically infinite amount of 3d scenes can lead to
same 2d image
PROBLEM OF OBJECT INVARIANCE
Need to recognize 3d object from all possible 2d images
Constructing 3D
- Reconstruct scene from image
- Monocular depth cues
- Requires one eye only
- Can be represented in 2D art/pictures
- Binocular depth cues
- Requires both eyes
- Binocular disparity
- Slightly different images on each retina due to distance
between eyes - Distance information is provided from both
MONOCULAR DEPTH CUES
Occlusion: Closer objects occlude objects further away
Proximity to Horizon (Relative height): Distance from object origin to horizon. Closer proximity = further away
Vanishing Point
Ponzo illusion
Relative size: Similar objects appear smaller as they are further away
Relative Height
EBBINGHAUS ILLUSION
- Size constancy
- Size is partially determined by contex
FORCED PERSPECTIVE
Illusion only works when viewed from a single point/angle of perception.
BINOCULAR DEPTH CUES
Each eye gets a slightly different view of the visual
field (check it yourself)
* Binocular disparity
* For a given object, the degree of offset is determined
by how far away it is
MOTION ADAPTATION EFFECT
Neurons get fatigued staring at still image and it starts appearing to move. Apparent motion can even work with random dots – if you prime people to see it or “direct” the motion
Structure from motion
* Rotating objects provide information about 3d shape
* Ballerina illusion
MOTION PAREIDOLIA
Detecting pattern where there is no real pattern
MOTION PARALLAX
- Movement of scene when you are moving
- Close things move by faster
- Helps with depth perception
TEMPORAL RESOLUTION
- Ability of visual system to separate events over time
- Flicker fusion – can’t separate, “fuse” together
- Speed of sight
- ~ 30 ms (?)
- But not uniform!
- Dorsal vs. Ventral (temporal hierarchy)
our perception of the world lags behind by ~200 millisecond
VISIBLE PERSISTANCE
- Sustained activation after stimulus
- Allows for temporal integration of quickly presented stimuli
- Movies
- Animation
- Light “tracers”
MOTION BLUR
- In humans, due to visible persistence
- In film, due to the way a camera works
- Not explicitly present in animation!
