Object Perception Flashcards
There are _ theories of object recognition: _, _, _, and _.
4
Template matching models
Feature matching Models
Recognition-by-components
Configural models
Theories of Object Recognition
_ presents that we detect patterns by matching visual input with a set of templates stored in memory.
Template Matching
Theories of Object Recognition
_ uses a set of transformations to best alight the object with a template (using: _, _, and _.)
Tamplate Matching
Translation
Rotation
Scaling
Template Matching
A _ is a pattern, like a cookie cutter or a stencil. It can be used to compare individual items to a standard.
Template
Theories of Object Recognition
_ detect objects by the presence of features.
Feature Theories
Theories of Object Recognition
_ search for simple but characteristic features of an object; their presence signals a match.
Feature-matching Models
Feature Theories
The feature-matching approach also lends itself well to the idea that processing of information in the brain is _.
Parallel
Feature Theories
Feature Theories need to also know how the features related to each other, also known as _.
Structural theories
Theories of Object Recognition
The _ model provides a possible method for recognizing three-dimensional objects across variations in viewpoint or exemplars (Biederman, 1987). (p. 78).
recognition-by-components (RBC)
RBC Model
The current model proposes that a set of _ geometrical three-dimensional shapes, such as cylinders and cones, can be used to represent just about any object; in the language of the model, these shapes are called _.
24
geons
Laban.
Laban!
RBC Model
Choice of shape vocabulary seems a bit _. However, choice of geons was based on non-accidental properties. The same geon can be recognized across a variety of different perspectives:
Arbitrary
RBC Model
_ is possible except for a few accidental viewpoints, where geons cannot be uniquely identified
Viewpoint Invariance
RBC Model
_ is easier when geons can be recovered.
Prediction
RBC Model
_ disrupts geon processing more than just deleting parts of lines
Disrupting vertices
RBC Model
In general, the effect of _ occurs when a stimulus or task facilitates processing a subsequent stimulus or task
priming
RBC Models
Theory does not say how _, _, and _ details are processed. These are often important to tell apart specific exemplars or similar objects.
color
texture
small
RBC Model
What are the problems of the RBC Model
- Structural description not enough, also need metric info to distinguish between specific exemplars or similar objects
- Difficult to extract geons from real images
- For some objects, deriving a structural representation can be difficult
Theories of Object Recognition
_ states that individual instances are not stored but what is stored is a representative element of a category.
Configural Models of Recognition
Theories of Object Recognition
_ is where recognition is based on “distance” between perceived item and prototype.
Configural Models of Recognition
Theories of Object Recognition
_ have been successful in the domain of face recognition.
Configuration Models
Configural Models
_ is a representative element of a category.
Exemplar
Configuration Models
Specific faces are described by their _, as defined by quantified average proportions in a population.
deviations from the prototypical face
Configuration Models
All faces would have the same component parts in the same spatial arrangement, but their _ and _ make each unique.
relative sizes
distances
Configuration Models
We create a caricatured version of the exemplar by _.
Moving away from the norm
Configuration Models
By disrupting _, it becomes easier to process the individual parts
holistic (configural) processing
Configuration Models
_ states that configural effects often disappear when stimulus is inverted.
Face Inversion
Top-down and Context Effects in Object Recognition
_ can often help in identification of an object but can also alter the interpretation of an object.
Context
Context Effects
_ states discriminating between letters is easier in the context of a word than as letters alone or in the context of a nonword striking.
Word Superiority Bias
Context Effects
_ suggests that information at the word level might affect interpretation at the letter level.
Word Superiority Effect
Context Effects
_ is a neural network model for how different information processing levels interact.
Interactive Activation Model
Interactive Activation Model
Levels interact in _ as how letter combine to form words (feature to word level) and _ as how words affect detectability of letters (word to letter level).
Bottom up
Top-down
Interactive Activation Model
There are three levels of interactive activation models:
feature
letter
word
Interactive Activation Models
_ represent features, letters and words where each has an activation level.
Nodes
Interactive Activation Model
Connections between nodes are _ or _.
Excitatory
Inhibitory
Interactive Activation Model
Activation flows from _ to _ to _ level, and back to _ level.
feature
letter
word
letter
What method is being described in these statements:
1. Acquire brain data for different stimuli (e.g. bottles and shoes)
2. Train a classifier (such as the neural network on right) to discriminate between bottle voxel patterns and shoe voxel patterns
3. Test classifier on novel images
Pattern Classification Model
Pattern Classification Model
Haxby et al. (2001)
can predict with 96% accuracy stimuli from 8 categories, namely:
Faces
Cats
Scissors
Chairs
Houses
Bottles
Shoes
Scrambled Pictures
What is being described in the statement: Engel & co-workers (e.g 1992) suggested that neurons representing the same object fire in synchrony.
Reconstructing Mental Images
