Notes: Vision

PDF

Where we are going:
    How is a distorted and upside-down 2-D retinal image transformed into the 3-D world we perceive?

Light

No species can see in the dark, but some are capable of seeing when there is little light
Light can be thought of as
 - Particles of energy (photons)
 - Waves of electromagnetic radiation (has a wavelength)
Humans see light between 380-760 nanometers in wavelength

Properties of light:

________________ – perception of color

________________ – perception of brightness


Source: http://www.perret-optic.ch/optometrie/Vision_des_couleurs/vis-couleur_gb.htm

The Eye

Focusing an image

______________ - contractions of ciliary muscles to deform the lens and change the focus

Retina

Transduction

transduction - conversion of one form of energy into another

visual transduction - turning _____________ into a _____________ _____________

how does this happen: pigment absorb photons and react

Rods & Cones

________________________ - cells specialized for visual transduction

rods - specialized for seeing ________________________
    more sensitive to photons than cones
    signals from many rods are pooled into one retinal ganglion cell

cones - specialized for seeing________________________ (more later)
    in most humans, there are 3 different cones sensitive to 3 different wavelengths of light



Rhodopsin - a ______________ that changes shape when it absorbs ______________
    (you won't be tested on cyclic GMP, just rhodopsin in general:)
    cyclic GMP keeps sodium channels open
    when rhodopsin absorbs light, it breaks up cyclic GMP

    when light hits rhodopsin, this:
        increases/decreases the amount of Na+ entering the cell
        depolarizes/hyperpolarizes the cell
        increases/decreases glutamate release

Rhodopsin: Source: wikipedia.org/wiki/Rhodopsin

_________________ - the ability to see when light is dim, requires _________________ photoreceptors

_________________ - the ability to see details (resolution), requires ________________ photoreceptors



_________________ - the center of the retina
    where the lens focuses the image
    a high/low concentration of cones
    a high/low concentration of rods

_________________ - the area of the retina where the axons from the retinal ganglion cells leave the eye

Trick for seeing in the dark - don't look directly at what you want to see.
    Why does this work?

Visual pathway

Source: http://www.dgward.com/physo101/sm06_pages/labs/Peripheral%20Vision%20and%20Visual%20Pathways.htm

Lateral geniculate nucleus (LGN)
    - part of the thalamus, which is a relay station between most sense systems and the cortex
    - exact role is unclear
    - maybe involved in: making visual information more efficient, focusing attention, saccades

Visual cortex (more later)
    - performs the processing on visual information to allow us to perceive visual scenes/stimuli

Information from LEFT visual field goes to RIGHT visual cortex (and vice versa)
    NOT left EYE to right visual cortex

Retinotopic mapping
    - If two retinal ganglion cells that are close together in the retina, their axons end close together in the visual cortex
    - The retina is "mapped on" to the cortex

Low-level Visual Processing

"low-level" refers to early in the visual pathway & dealing with simple visual stimuli like brightness, edges & color.
"high-level" refers be areas that receive the pre-processed information from low-level parts of the visual system and that process more advanced stimuli like motion, faces, object-recognition & visual space

Receptive Fields

Definition: The area of visual space that stimulates or inhibits a neuron (or neural tissue)

The stimulus might be simple or complex. Examples:
  - Some neurons might be stimulated by any light in a precise spot in the top right corner of the visual field
  - Some neurons might be stimulated by a vertical edge anywhere on the left
  - Some neurons might be stimulated by faces anywhere in the visual field

Receptive fields become larger
    farther away from the fovea - don't need to know exact location, just want to notice something
    at higher levels of the visual system - just want to react to a face (for example), doesn't matter where it is

Hubel & Weisel

Videos: Intro & long version

Color

In most humans, there are three kinds of cones
    - each with a different photo-sensitive pigment called iodopsins
    - each of the three iodopsins is sensitive to different wavelengths of light

Number of cones varies
    - some animals & people (with color blindness) have only 2 kinds of cones
    - some animals (birds in particular) have 4 kinds of cones

High-level Visual Processing

Sensation vs Perception

Incoming signal

    - left/right visual fields to different hemispheres
    - each hemisphere contains intermingling cells that respond to left and right eye
    - retinotopically organized
    - no perceptual processing (still basically "pixels")

Visual cortex

Source: http://www.techcyn.com/feature.php?id=f2&issue=1

______________ cortex - stripped appearance

primary visual cortex (striate) -> parastriate -> prestriate/extrastriate -> temporal lobe
increasingly complex processing

Dorsal vs Ventral stream

Ventral/what stream     - Used for recognition/identification     - Captures fine details but is slow     - Conscious awareness and interactions with long term memories.     - Notices fine details (fovea)     - Neurons respond to objects anywhere in the visual field

Dorsal/what stream     - Used for visually guided behavior     - Sensitive to motion, fast processing     - Not conscious processing     - Neurons respond based on where visual attention is allocated

Face recognition

Fusiform Face Area  - in ventral stream / temporal lobe  - somewhat right lateralized  - particularly active to faces, though somewhat to objects in general  - example of extreme specialization in the visual system

Principles of Visual Processing

• Parallel processing
• Hierarchical
• Functionally segregated

These principles apply to many neural systems besides vision