Reading...
Play button
Play button
Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
33 Cards in this Set
- Front
- Back
|
what are the strengths and limitations of Gestalt's perspective?
|
Strengths- Relied on heuristics, and described the heuristics. Two lines close to one another signify liklihoo dof an object.
limitations: Doesn't explain the mechanisms. Gestalt perspective does not explain how our mind does this. |
|
|
object recognition is hard. Why?
|
Object recognition is hard because the stimulus is underspecified. There is always an alternative for what you are seeing, touching, hearing etc, experiencing.
|
|
|
Describe template theories of object recognition:
|
Earliest theories. Our minds have templates of every object and the objects we see are matched up to the templates.
|
|
|
Describe feature theories of object recognition:
|
We recognize different features that make up objects.
|
|
|
Name of two models to do with feature theories
|
Pandemonium, interactive activation.
|
|
|
example of feature theories:
|
Everyones handwriting is different. If 100 people write the letter 'A' it is going to look different in every case, however in the majority we will still be able to see that it is an A.
|
|
|
Pandemonium model stages:
|
Image demon, feature demons, cognitive demon, decision demon.
|
|
|
what do the demons in the Pandemonium model stand for:
|
nodes
|
|
|
Who is responsible for the Pandemonium model? and year?
|
Selfridge 1959
|
|
|
What are 3 stregths of Pandemonium model?
|
1. Each node only has one job to do.
2. includes parallel processing. 3.Neurologically sound |
|
|
what kind of process is Pandemonium model?
|
Bottom up. Data Driven.
|
|
|
Limitiations of the pandemonium model:
|
1. Where is the conceptually driven, top down processing?
2. doesnt explain how to go from letters to words. |
|
|
Who is responsible for the interactive activation model
|
Mcllelhand and Rumelhart
|
|
|
How many levels are there as part of the interactive activation model?
|
3
|
|
|
What are the 3 levels of the interactive activation model?
|
input units
hidden units output units |
|
|
How many input units are there in the interactive activation model?
|
9
|
|
|
What is connectionsim? (feature of the IAM)
|
When all the nodes are connected together.
|
|
|
what are positive and negative weighting?
|
when input units, have weighting either positive for similarity to the word units, or negative to completely different.
|
|
|
What is preactivation?
|
Mind pre-activates a likely word. Example "the Bear climbed the,....." likely to pre activate tree.
|
|
|
How does IAM involve both types of processing?
|
Activation through the levels as per pandemonium- data driven, however word recognition feeds down the other way as well - Conceptually driven.
|
|
|
How does a neural network training work?
|
Neural pathways are either strengthed or weakened as time goes on and it is evident which pathways are working and which arent. CAn be used for word recognition or motor actions- such as an infant trying to grab an object.
|
|
|
What are Geons?
|
letters of the object alphabet.
|
|
|
What theory did Beiderman assoc with?
|
Recognition by components. All objects are made up of Geons.
|
|
|
What are feature detectors?
|
Way geons can be identified.. Areas of intersection and edges of objects
|
|
|
How many Geons are there?
|
36
|
|
|
What 3 things are special about Geons?
|
1. Viewpoint invariance
2. Resistance to visual noise 3. Discriminability |
|
|
What is viewpoint invariance?
|
Geons are easy to identify form any viewpoint or angle.
|
|
|
Resistance to visual noise?
|
If part is covered for eg still can identify.
|
|
|
Discriminability
|
no two Geons look the same.
|
|
|
limitations of the Geon/recognition by components/Beiderman model:
|
Data driven bottom up processing. Which is fine but sometimes we see the whole picture before the shapes.Ignores parallel processing.
|
|
|
What did Navon show us?
|
Study using letters arranged in shapes. Study showed us that we sometimes see wholes before parts.
|
|
|
What type of processing happens in the left hemisphere?
|
Local processing
|
|
|
What type of processing happens in the right hemisphere?
|
Global processing
|
