1/42
Looks like no tags are added yet.
Name | Mastery | Learn | Test | Matching | Spaced | Call with Kai | Chat |
|---|
No analytics yet
Send a link to your students to track their progress
Why is object recognition is important?
Necessary for applying knowlede and learning
What is like for patients with associative Agnosia?
Can see objects but unable to link knowledge to what the object is
Where does recognition begins with?
Features
The advantages of a feature-based system?
-Building Blocks of preception
-Objects commonalities needed for search
-Cruical for visual search
The differences between parallel and serial search
Parallel: pop out effect--one feature at the time
Serial: One feature need, which is need to look one by one in a group or series
Apperceptive agnosia
the lack of ability to organize elements that they see in regards to an intact object.
What is Tachistoscope Methodology process?
1. A stimulus (word, letter, or random set of letters) flashed
2. Identification of the stimulus by the participant
3. The experimenter can stimuli, presentation time, or the context
Frequency effects
The effect of recognition after presentations for printed words compared to words appearing only rarely
Priming Effects
-recently viewed words = recognized more often
-repetition priming
What is a neural network?
A neural network is made up of groups of neurons that work together to detect different things.
What is a receptive field in a neural network?
A receptive field means a detector responds to a specific shape, pattern, or location.
What happens to a detector in a neural network as it receives more input?
The more input a detector receives, the more active it becomes.
What occurs when a detector's activation reaches a certain threshold?
The detector fires.
What effect does one detector firing have on other connected detectors?
It increases the activation of the other detectors it is connected to.
How do simple detectors relate to complex detectors in a neural network?
Simple detectors connect to more complex detectors, allowing the network to recognize more complicated things.
What could explain feature nets in two words?
-Reptition priming
-Familarity
reptition priming
more recent words are more recognized
Familarity
more frequent words in the langauge are more recognized
What happens to the detectors when both of these results?
detectors have a higher activation level

Word-Superiority effect
tendency to identify a letter more accurately when it is part of a word than when it is presented by itself
What is world superiority based on?
Probability
The latter of bigram detectors
Feature -> letter -> Bigram -> Word
word superiority effect: Feed Foward
the letters only go one way
Feature net model have what
-inhibitory and feedback
Inhibition
Helps singal the reduction activation
Feedback
Helps with the detection effect activation of lower order detector
Feature net featues
-Not represented local knowledge
-contains distrbuation of knowledge
-efficiency exist for the sacrifice accuracy
How can errors aise from feature net?
The limitations of the network's ability to deal with inputs and recover fromerros
Bottom Up Perception features
senses and feature
Bottom up perception: information and perception
-Parts are organized into a whole
-stimulus provides information
Top-Down Perception features
Context and priming
Top Down perception: Information and perception
-the whole and THEN, the indivdual parts are need
-experience driven > stimulus data
-a source of misperception
TRUE/FALSE: obeject is complicated
TRUE
Recognition by Geons
-theory of object recognition is based on genos
-feature nets that 'bulid up' of geons to get to a whole objects
What are geons?
basic units of objects
The Geon viewpoint invariant recognition
the recogniation of objects from just a few geons
Theoy of recognition by mutiple views
the ability to identification of an object from multiple angles and views
A disadvantage of Theoy of recognition by mutiple views
it is just too many views to store objects
A solution to the disvantage
-a subset
-a mental images to rotate to store views
What happens after the solution is applied to the Theoy of recognition by mutiple views?
-Speed of recognition
-Rotation takes time to set
Prosopagnosia
inability to recognize different faces
What makes special identification occur?
-interpretation of face
-high expertise
-recognization of special individuals
-important configurations of component parts
Hollistic Processing
perceptual analysis that emphasizes overall shape of object
-face perception reflects the whole