object perception

Information required to identify distal stimulus

• light sources

• Reflectance of surfaces

• Surface orientation

• Viewing position

Light sources

primary (sun, light bulbs), secondary (moon, walls)

reflectance of surfaces

proportion of light reflected; absorption spectra: surfaces absorb certain wavelengths

Surface orientation

position of surface causes different amount of light to fall on it/reflect off it

viewing position

scene appears different from different viewing angles

Inverse projection problem

how do you go from the proximal stimulus to identifying the distal stimulus?

• “uncreate” the image, using reverse geometry?

• retinal image is 2-D, real world is 3-D

• any 2-D image may have many potential 3-D sources that created it (underdetermination)

• but the visual system selects the proper interpretation (most of the time)

• how does it do this?

Contour detection

• contour

• Ganzfeld

Contour

border between light and dark areas in an image

Ganzfeld (“total field”)

• sort of sensory deprivation; developed to test for telepathic communication

• translucent ping-pong ball halves placed over eyes; headphones deliver white noise

• causes some observers to “blank out” or hallucinate

• sharp luminance contours must be important to visual system

Dynamic stimuli: microsaccades

minute, involuntary, almost imperceptible eye movements that occur during fixations; purpose seems to be to change the position of the image on the retina

• Pritchard (1961):

  • observers wore contact lenses with attached device

  • projected constant image onto retina

  • image disappeared within ~3 seconds

  • eye responds best to changing stimuli; static pattern will quickly fade from view

Gestalt Psychology: basis

motion could be produced by a succession of stationary stimuli (Max Wertheimer, 1912)

Gestalt Psychology: motto

“the whole is different than the sum of its parts”

Gestalt Psychology: illusory contours

contours not present on retina, but still affect contour perception

• e.g., Kanizsa triangle (1979):

• (end-stopped cells are activated by illusory contours)

Figure-ground segregation

Generally, figures:

• appear to be in front

• are smaller

• have well-defined shape

• are meaningful

• have more detail

• differ from background in lightness

• e.g., Rubin figure (1915)

Laws of Perceptual Organization

• proximity/nearness

• Similarity

• Good continuation

• Closure

• Common fate

• Meaningfulness/familiarity

• Law of Prägnanz

Proximity/nearness

things near to each other tend to be grouped together

Similarity

similar things tend to be grouped together

Good continuation

points that, if connected, would result in either straight or smoothly curving lines, tend to be seen as belonging together; and lines tend to follow the smoothest path

Closure

A space enclosed by a contour (real or illusory) tends to appear as a figure

Common fate

things that are moving in the same direction tend to be grouped together

Meaningfulness/familiarity

things that are meaningful or familiar tend to form groups

Law of Prägnanz (“good figure” or “simplicity”)

every stimulus pattern is seen in such a way that the resulting structure is as simple as possible

Gestalt Pros & Cons

• holistic approach

• Operates post hoc, not good for predictions

• Not good at explanations

• Vague definitions (“simplicity”)

Recognition

perceiving something as previously experienced

• e.g., Have you seen this before?

Identification

naming or classifying an object

• e.g., What is it? What is it used for?

Object recognition and identification: Template theories

• compare input to a model or “template” stored in memory

• stimulus categorized by exact match

Object recognition and identification: pros and cons

• successfully used by machines

  • e.g., MICR numbers at the bottom of a cheque

• intolerant to __________ in a stimulus

• too many templates required

• cannot account for novel stimuli

• cannot handle _______: