object recogniton Flashcards

1
Q

What is perception?

A

The ability to extract meaning from sensory input, involving audition, taste, touch, and smell, but predominantly studied in vision.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What percentage of neurons in the cortex are involved in vision?

A

Over 50%.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What are the two processing streams in vision, and what do they represent?

A

The ventral stream is the “what” pathway (object recognition), and the dorsal stream is the “where” pathway (spatial location).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What does the study by Tootell et al. (1982) suggest about the visual system?

A

It supports the idea that a near-perfect representation of the external world is projected onto the primary visual cortex.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

How was the study by Tootell et al. conducted?

A

Monkeys’ eyes were held open, exposed to patterns, and injected with radioactive isotope glucose while unconscious.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What are the three stages of the object recognition model?

A
  1. Local features (e.g., edge detection, contrast).
  2. Shape representation (e.g., Gestalt principles, feature integration).
  3. Object representation (stored knowledge).
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What is the main idea of Gestalt principles?

A

The whole visual percept is more than the sum of its parts, as the system makes assumptions based on the environment and prior experience.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What are examples of Gestalt laws of perceptual organization?

A

Good continuation and closure, which can create illusions like illusory contours.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

What is a ‘primal sketch’ in shape perception?

A

A representation of primitives like edges, orientations, and colors created through bottom-up processes (Marr, 1982).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

How do top-down processes influence shape perception?

A

They group primitives into larger structures like lines, curves, and blobs using Gestalt laws.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What are the three models of object recognition?

A

Template Matching (prototype theory).
Feature Analysis.
Recognition by Components. (Structural Theory)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

What is the core idea behind template matching?

A

Visual input is matched against internal templates or representations stored in memory.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Where is template matching used outside human perception?

A

Computer-based object recognition programs.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What does feature analysis propose about object recognition?

A

It begins with lower-level analysis, searching for simple but characteristic features of an object.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

What neurological evidence supports feature analysis?

A

Orientation-selective cells in the visual cortex.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

What is recognition by components (RBC)?

A

A model where objects are represented as arrangements of simple 3D shapes called geons (Biederman, 1987).

17
Q

What makes geons crucial in recognition by components?

A

They are viewpoint-invariant and recoverable from 2D retinal images.

18
Q

What happens when geons are made non-recoverable?

A

Object recognition is impaired.

19
Q

How can color affect object recognition?

A

It can aid recognition by providing additional visual cues.

20
Q

Is object recognition a passive or active process?

A

It is a constructive process that combines bottom-up and top-down processing

21
Q

Are the models of object recognition mutually exclusive?

A

No, object recognition may involve elements of template matching, feature analysis, and recognition by components, depending on the situation.