Visual perception I: Lower-level vision Flashcards
Distal Sensory Processing
Perception based on gathering remote information and processing it in the brain. E.g. Vision is based on distal stimulus in the form of light entering our eyes.
The opposite would be proximal sensory processing, E.g. touch.
How is light turned into an image in the back of the retina?
The light is refracted 80% by the cornea and 20% by the lens. This creates an inverted image in the back of the retina. The inversion is reversed later in the visual processing, the important thing is that the proportions of the image is not distorted.
What did Yarbus show in his initial eye tracking studies?
He showed that, rather than our eyes resting passively on a visual stimulus, our eyes actively examine the stimulus. He showed that eye movements consist mainly of fixations (focusing on specific point) and saccades (moving from one fixation to the next). Furthermore, the eye movements are focused on the most important parts of the visual stimuli, e.g. when seeing a face, there are most fixations at the eyes and the mouth, because they convey the most information.
Pupil
A hole in the iris which can be expanded or contracted. This functions as a light filter, where the pupil will dilate in dim light settings to let more light in, and contract in bright settings to let less light in. If the light levels change quickly from low to high, the pupil rapidly contracts to protect the eye.
This is not the only way the eye compensates for light intensity, as most of the adaptation happens in the retinal photoreceptors.
Pigment Epithelium
Black layer in the back of the retina, which stops light from reflecting back into the eye. after having reached the photoreceptors.
Fovea
The fovea is a small pit in the retina where we have the highest acuity of seeing colored structures in images. When you try your best to focus on something, the eye is aligning itself so that the light hits the fovea. Besides the fovea, rods are the dominating photoreceptor, taking care of most peripheral vision.
The fovea eye pit does not have any rods or other neurons, only millions of tightly packed cones. By grouping en masse, cones get optimal exposure to soak up light as it comes into the eye, allowing them to create the sharpest possible image.
The fovea function also includes the discernment of other image details, such as distinguishing between different colors and sensing three-dimensional depth.
Photoreceptors + Convergence
Rods and Cones located at the retina. (See dedicated cards)
Both rods and cones are nerve cells which give a graded output to bipolar cells. The bipolar cells also give a graded output on to the ganglion cells, which converts these into action potentials. There are far fewer ganglion cells than photoreceptors, which shows that already in the retina, the visual information is compressed, and that the photoreceptors and bipolar cells help with convergence of the visual image.
Rods
Rod photoreceptors are sensitive in dimly-lit environments, and assist the eye in night vision and seeing in black and white. These photoreceptors contain a protein called rhodopsin (also called visual purple) that provide the eye with pigmentation in low-light conditions.
Cones
Cone photoreceptors are activated by bright lighting and help the eye to see color. This type of photoreceptor contains proteins called photopsins (or cone opsins) that help create color pigments for the eye to view.
There are three types of cones.
Short-wave cones (also called blue-light cones)
Middle-wave cones (also called green-light cones)
Long-wave cones (also called red-light cones)
The cones have been assigned a colour, but it is more accurate to think of it in terms of a wavelength sensitivity based on a distribution on the spectral field. e.g the ‘red’-cone does not have peak sensitivity in the red area of the colour spectrum, but it is the only cone type which sensitivity distribution reaches into the red colour. The colour output is based on the relationship between these three cone types.
Horizontal and Amacrine Cells
Cells that perform so-called horizontal processing of visual stimuli. Horizontal cells are located between the photorecptors and the bipolar cells, and Amacrine cells are located between the bipolar cells and the ganglion cells.
They generally help with contrast detection and generally differences between neighbouring photoreceptors.
Three Levels of Light Vision
- Low light: Skotopic
- driven by rod photoreceptors with reduced sensitivity to red light
- Medium light: Mesotopic
- driven by a mix
- High light: Photopic
- largely driven by cone photoreceptors
Purkinje Shift
The Purkinje shift is a shift in perceived brightness of colours when going from light conditions to low-light conditions. Due to the difference in wavelength sensitivity between rods and cones, high wavelength colours (Red and Yellow) appear bright in photopic vision(Light), whereas low wavelength colours (blue and green) appear bright in skotopic vision (Low-light).
Optic nerve
The axons of the ganglion cells throughout the retina come together to form the optic nerve.
Blind spot in the eye
Where the optic nerve leaves the eye (15 degrees off the fovea), there is a blind spot, because no photoreceptors are present. What’s interesting is that even though nothing is perceived in this spot, it is filled in by what the brain deems most likely based on the surroundings.
Ganglion Cells
Ganglion cells get input from photoreceptors through the bipolar cells, and turn this information into action potentials. The axons of ganglion cells make up the optic nerve.
Each ganglion cell has its own receptive field based on the location of the photoreceptors in the retina that signals to the ganglion cell.
Haynes: “The receptive field is the region of visual space where a suitable* stimulus can directly cause an increase or decrease in spike rate”
*suitable in the sense that the stimulus may still need additional features such as the right rotation in order to alter spike rate
On- and off-center ganglion cells
The major functional subdivision of ganglion cells in the mammalian retina is into ON- and OFF-center ganglion cells. ON-center cells are depolarized by illumination of their receptive field center (RFC), while OFF-center cells are depolarized by decreased illumination of their RFC. It is commonly assumed that ON-center ganglion cells receive excitatory input from ON-cone bipolar cells, while OFF-ganglion cells are excited by OFF-cone bipolar cells.
The center activation and the off-center activation is combined to determine the spike rate of the ganglion cell.
Mexican hat filter
The Mexican hat filter is a 3d representation of the on-off activation in a single single ganglion cell. The filter has a high spike in the middle and is lower around the spike, mimicking the ganglion cell response pattern.
Binocular visual field
The part of the visual field which is seen by both eyes. Seeing with both eyes allows depth processing.
The opposite of Binocular visual field is: Monocular visual field
Left and right Visual Hemifield
The left and right visual hemifields are the visual field to the left and right of our fixation point. Both eyes are involved in processing both hemifields, however the ipsilateral eye corresponds to a larger area. The optic nerve projects information from both hemifields until it reaches the optic chiasm, which separates the axons coming from the right and left hemifields, and directs them to the contralateral side, making up the right and left optic tract. The visual system functions via contralateral processing, meaning that the left hemifield is processed in the right part of the brain, and vice versa.
Lateral geniculate nucleus
The lateral geniculate nucleus (LGN) is the thalamic relay station to the visual cortex. It is a small mass on the side of the thalamus, where the axons from the optic tracts synapse. Visual information from the LGN goes directly to V1 in the visual cortex.
The lateral geniculate nucleus has 6 layers that are bended in a knee shape. The six layers can be categorized into magnocellular layers (1 and 2) and parvocellular layers (3, 4, 5, 6), their functions are described in the next cards.
V1, striate cortex, primary visual cortex
All three names of the same area in the visual cortex.
V1/primary visual cortex because it is the first cortical structure involved in the visual processing system.
Striate cortex because it has a stripe. Striate means ‘striped’. The white stripe is seen when cutting into the cortex, where the input layer for visual processing is so heavily myelinated that it is visible to the human eye. This is helpful to distinguish the primary visual cortex from other parts of the brain.
V2 does not have this stripe.
Hubel and Wiesel
Sigh..
Tuning width of V1 cells
The tuning width is the degrees in which a particular cell will elicit a signal, based on the orientation of a line. If a cell responds most strongly to lines a a 30 degree angle, and has a tuning width of 40, it will also show a small response to lines orientated at 10 and 50 degrees.
The median tuning width is 41.5 degrees, but it varies from 0-180 degrees.
