3. Perception Flashcards
distal stimulus vs. proximal stimulus vs. percept
distal stimulus: actual stimulus itself
proximal stimulus: the information
consisting of the reception of the distal stimulus
percept: meaningful interpretation of the proximal stimulus
What is one fact that supports distinguishing proximal stimuli from percepts
size constancy
what is form perception (gestalt)
distinguishing of the display into objects and background
what do subjective contours mean in gestalt psychology for perception
when we simplify a complex display into a simple one, by sometimes imagining edges that arent actually there
What some (5) of the Gestalt principles of perceptual organization
Principle of proximity
Principle of similarity
Principle of good continuation
Principle of closure
Principle of common fate
Explain top-down vs bottom-up processes
Top-down: Perceiver’s expectations/theories guide the pattern recognition process
Bottom-up: Perceiver starts with small bits of information and combines it to form a percept
What are 3 bottom-up models of perception
Template matching
Featural analysis
Prototype matching
Describe template matching
Type of bottom up perception model
- we read in patterns and compare them to previously stored patterns (templates)
What does the template matching model fail to explain
- We will need to store an impossibly large number of templates
- How do templates get created and how they are kept track of
- Patterns are recognized to be the same even if their stimulus patterns differ (etc. blurriness)
Describe featural analysis
Bottom up perception model
We conduct perception by searching and recognizing features, and then recognizing the object as a whole
What are some neurophysiological experiments (3) that support featural analysis
Implanting electrodes to the retinas of frogs: noticed that “edge” and “bug” stimuli caused certain cells to fire more frequently
Hubel and Wiesel found evidence of separate horizontal + vertical line detectors in cats and monkeys
Visual search: participants recognize letters by features
What was the finding of Neisser’s ZQ visual search task
Participants took longer picking out Z or Q from a group of letters, depending on how similar the letters around them were
Showed that we identified letters by features (featural analysis model)
What is Selfridge’s Pandemonium model for feature analysis
Think of a neural network, but instead of feature maps you have feature demons on each layer
And instead of weights you have the volume of demon’s screaming as they communicate between layers
What does the feature analysis model fail to explain
What is a feature, and how is this decided?
since an observer would need to have a list before starting perception, and this list can be arbitrarily large, how do we do perception quickly?
Describe prototype matching
We match input to a stored prototype: its idealized representation