PA 3 Flashcards
2D pattern matching theory. Template theory
what is it
what are its problems
- multiple templates held in memory
- Compare stimuli to templates in memory for one with greatest overlap until match is found
- Problem of imperfect matches
- Doesn’t account for flexibility
2D pattern matching theory. Prototype theory
Evidence for
Possesses the average of each individual characteristic
- No match is perfect
Present object based on prototype, prototype not shown. Yet p.p are confident they have seen prototype.
Feature theories of 2D pattern matching
pattern consists of a set of characteristic features.
A = 2 straight lines connected by a bar.
BUT also need to know relationship between features
2D pattern recognition. Structural descriptions
describe the nature of the components of a configuration and the structural arrangement of these parts
Capital letter T = two lines. bottom supports top. bottom bisects top. … more detail of structure
3D object recognition
- first separate object from background.
- Must be processed to give a description. which can then be matched to descriptions of object in ur memory
Marr & Nishihara’s theory of 3D object recognition. cylinders
objects comprised of cylinders - must specify relationship between cylinders.
Hierarchical organisation of cylinders
Biederman’s GEON theory
alternative to cylinder theory.
there are 36 diff shapes “GEONS” that all objects can be broken down into.
Concavity helpful in segmenting visual images into parts.
Biederman support
Found some cortical neurons sensitive to GEONS in monkeys
- Assessed response of individual neurons in the inferior temporal cortex to change in geon or change in size of object
- Some neurons respond more to geon changes, providing support for geons
Evaluation of Biederman
- flexible and comprehensive system for describing objects
- Experimental models don’t provide critical test.
- Doesn’t explain how descriptions are matched to those stored
- ignores important factors. e.g. colour, top down processing
viewpoint dependant theory
assume changes in viewpoint reduce the speed/and or accuracy of object recognition.
What are key challenges in object recognition
- Intra-class variation – Objects within a category vary in appearance (e.g., different types of chairs).
- Viewpoint variation – Objects appear different from various angles.
- Partial occlusion – Objects can be partially hidden but still recognized.
What is the difference between viewpoint-dependent and viewpoint-invariant theories?
- Viewpoint-dependent theory – Recognition depends on the viewing angle (Tarr, 1995).
- Viewpoint-invariant theory – Objects are recognized regardless of the viewpoint (Biederman, 1987).
What are the four main theories of 2D pattern matching?
- Template theory – We store exact copies (templates) of patterns in memory.
- Prototype theory – We store average representations of patterns.
- Feature theory – Objects are recognized based on distinct features (e.g., the letter “A” = 2 diagonal lines + crossbar).
- Structural description theory – Objects are recognized by their parts and spatial relationships (e.g., a “T” has a vertical and horizontal bar).
When is viewpoint-dependent recognition used?
Complex within-category distinctions (e.g., identifying different car models)
When is viewpoint-invariant recognition used?
- Easy categorical distinctions (e.g., recognizing a dog vs. a chair).
What are the key stages of object recognition?
- Structural description – Identifying object shape.
- Semantic representation – Linking the object to stored knowledge.
- Name representation – Retrieving the object’s name.
What is the difference between the Serial Stage Model and the Cascade Model of object naming?
- Serial Stage Model – Object recognition follows a strict step-by-step process (structural → semantic → name).
- Cascade Model (Humphreys et al., 1988) – Stages interact, allowing later processes (e.g., naming) to begin before earlier ones (e.g., structural description) are complete.
What is agnosia?
A disorder where a person cannot recognize objects, despite intact vision, memory, and intelligence.
What are the two main types of agnosia?
- Apperceptive agnosia – Early-stage deficit; problems with shape extraction (e.g., difficulty recognizing objects from unusual angles).
- Associative agnosia – Later-stage deficit; visual recognition is intact, but meaning cannot be accessed (e.g., patient can copy a drawing but doesn’t know what it is
What evidence supports agnosia as a distinct cognitive impairment
Patients with apperceptive agnosia struggle with unusual views of objects.
- Patients with associative agnosia can describe objects but cannot name or use them correctly.
