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;
50 Cards in this Set
- Front
- Back
|
Object perception
|
The ability to see an object as whole. Sometimes it takes place so quickly that we don't notice it and it organizes information in the image.
|
|
|
What are the goals of object recognition?
|
1. We need to be able to recognize an object and know what it is.
2. It is the basis of intelligent visual behaviour. 3. We describe the object mentally, record image of object and store it in long term memory. 4. We don't need a name for something to have object recognition, all you need is a feeling of familiarity. |
|
|
How does visual-motor coordination work?
|
In order to coordinate body properly you need to represent the object as a whole of its parts – can even do this without a name for it or without a memory representation of something
|
|
|
Top Down Theories
|
Top down theories involve using your knowledge, past experience, and beliefs to interpret what's in the image. When you have previous knowledge it can help you perceive objects - if you know what to expect it's easier to see it.
|
|
|
Perceptual set
|
You see something based on an image you've seen before.
|
|
|
What are some disadvantages of top-down effects
|
• Seeing only what you want to see/expect to see can be dangerous – for example if you don’t expect to see a lion it doesn’t mean that it isn’t there.
|
|
|
Bottom-up Theories
|
Bottom-up theories involve figuring out information straight from the sensors.
|
|
|
What are the three different stages of bottom-up theories?
|
1. Low level vision (pre-attentive analyses)
2. Visual routines (analyses requiring spatial attention) 3. Visual cognition (object recognition) |
|
|
Spatially parallel process
|
Things occur at every location of the visual field at the same time. It does not require attention and happens automatically.
|
|
|
Spatial serial process
|
Spatial serial process is really good at processing one part of the image but that means that the other parts of the image don't get much attention.
It's is essentially a spotlight focusing on one thing but does this at the expense of everything else. |
|
|
What is an example of spatial serial processing?
|
People drive straight into things that they are looking right in front of them because some parts of object recognition are serial – we have an attentional focus that isn’t where our eyes are foveated.
|
|
|
What happens without discontinuities?
|
Ganzfeld
|
|
|
Ganzfeld
|
when you put ping-pong balls over eyes and all you see is solid white/grey/black but when you do this, the eyes start to malfunction and you begin to hallucinate little flashes of colour. After a while the eyes move around independently by crossing and uncrossing the eyes without us noticing because we need landmarks in the image so we know where the eyes are pointing.
|
|
|
What happens without change?
|
Stabilized image technique
|
|
|
Stabilized Image Technique
|
We have an image on a contact lens and as the person’s eye moves so does the image which results in the same part of the retina always being stimulated. You see the image for a while but then it disappears if it doesn’t change position – this is why we don’t notice the blood vessels in our eyes.
|
|
|
How do you know an attribute is a feature?
|
By conducting a visual search task
|
|
|
Visual search task
|
Find the target (such as a horizontal line), manipulate the number of non-targets (distractors) and measure how much time it takes to decide whether the target is there or not.
|
|
|
Non-features
|
Finding a vertical black line is a lot harder when the distractors are black horizontal lines and vertical green lines (combination of attributes) – you see a pattern where your reaction time increases with the number of items in the display and increases when target is absent than when present.
|
|
|
Low Level Vision
|
It's spatially parallel and bottom up. The information is not taking into account expectations or memories, it’s just the image straight off the retina.
|
|
|
Visual Routines
|
• Spatially serial processing is similar to a spotlight. There’s one part of the image that gets a lot of analysis.
• You’re attending to something in that one part of the visual field at the expense of the other parts. • Attentional focus moves much faster because it doesn’t pass by intermediate points, just Point A point B. |
|
|
Visual Cognition
|
• Space is not relevant. Take the representation, aka symbolize information of what you’re seeing, use that information and compare it with a long-term memory representation.
• This involves naming or recognizing something as familiar. • Top down effects occur as a result of operations occurring in this final stage. In particular occur when your memory or expectations begin to drive what you see. |
|
|
Feature Analysis
|
Finding the discontinuity in an image (the edges) and categorize it. You deciding what kind of edge is - is it red, horizontal, blue, straight, curved, etc.
|
|
|
Pop-out
|
When targets and distractors provide feature. When doing the visual search task, we are able to ask if a target and distractor by feature. If so the target will pop-out despite the number of distractors.
|
|
|
Gestalt Grouping Principles
|
Gestalt believed the whole is different than the sum of the parts. We see more than just the image as we add into it.
|
|
|
Grouping by proximity
|
This is one of the simplest grouping principles. This occurs when this visual system groups disconnected contours together due to their distance.
|
|
|
Grouping by good continuation
|
This is grouping due to the tendency to see continuous lines.
|
|
|
Grouping by similarity
|
Grouping by how similar the features are.
|
|
|
Grouping by connectedness
|
Grouping depending on if contours are touching or connected.
|
|
|
What are some side effects of grouping?
|
1. Effortless texture segregation
2. Formation of illusory contours |
|
|
Effortless texture segregation
|
Textural element is when you have a pattern that is repeated. Segregation occurs when you pull apart and divide two things into sections.
Example: the letters L and T. They spacially similar as they are both horizontal lines with an intersecting vertical line. |
|
|
What are some examples of illusory contours?
|
Example: Necker’s cube. Here our visual systems sees co-linerarity and by good continuation its puts in a line that isn’t really there to explain why the circles are so perfectly lined up.
Example: Kanizsa triangle – we tend to see one triangle as being in front of the other. Image may be found on courselink. Example: Rubin’s vase can be interpreted two ways when you switch your contour background information. We either see a goblet or we see two faces on either side of the goblet. In this case you tend to the figure but tend to forget the background. |
|
|
How do we determine what is figure and what is ground?
|
We tend to see smaller things as the figure. High spatial frequency information tends to be seen as the figure.
|
|
|
Spatial serial analyses
|
This is analysis requiring spatial attention. We are building an object file by bringing all the characteristics of something together.
|
|
|
Spatial relations
|
We draw all the features of the objects together and talk about spatial relations between the parts. This includes left right top bottom inside outside, etc.
|
|
|
How do you know something is spatially serial?
|
In the visual search task what you’ll find is the response time to find the target increases with the number of items.
|
|
|
Illusory conjunctions
|
Illusory conjunctions occur when you bring together parts of different objects and form them into a whole. One image influences another. For example we may see the colour of one thing float over an influence how we perceive something else. It is essentially the miscombination of features.
|
|
|
What is an example of an illusory conjunction?
|
If someone is waving a gun in front of our face we won’t notice the robber’s face we’ll notice the gun. In this case the attributes of the other people in the bank tend to float over and influence how we perceived the robber. This happens because you weren’t properly paying attention to the robber and thus misremember.
|
|
|
Integrative amnesia
|
This causes the loss of the ability to make proper object files and thus these people have problems with spatial relations in particular. They may see a bunch of floating features but can’t make an object file as a whole. They couldn’t tell the difference between a bicycle and a pair of classes as both have circles and a bar.
|
|
|
What causes integrative amnesia?
|
Brain damage to the inferotemperol lobe.
|
|
|
Simultagnosia
|
Simultagnosia occurs if there are several objects in the same scene. These people are unable to keep two object files simultaneously and thus can only build on object file at a time and see one object.
|
|
|
Inverse projection problem
|
A particular image on the retina can be created by many different objects. If we know an object's shape, distance, and orientation we can determine the shape of the object's image on the retina.
|
|
|
What is it so difficult to design a perceiving machine?
|
1. Images on the retina are ambiguous
2. Objects can be hidden or blurred 3. Objects can look different from different viewpoints |
|
|
Viewpoint invariance
|
The ability to recognize an object seen from different viewpoints.
|
|
|
Law of pragnanz/law of good figure/law of simplicity
|
This is the central law of Gestalt psychology - every stimulus pattern is seen in such a way that the resulting structure is as simple as possible.
|
|
|
Gestalt grouping principles
|
1. Law of similarity
2. Law of good continuation 3. Law of proximity 4. Principle of common region 5. Principle of uniform connectedness 6. Principle of sychrony 7. Law of common fate 8. Law of familiarity |
|
|
Figure-ground segregation
|
When we see a seperate object it is usually seen as the figure that stands out from its background.
|
|
|
Properties of figure and ground
|
-figure is more “thinglike” and more memorable than the ground
-figure is seen as being in front of the ground -the ground is seen as unformed material and seems to extend behind the figure -the contour separating the figure from the ground appears to belong to the figure (border ownership: although figure and ground share a contour, the border is associated to the figure) |
|
|
Recognition-by-Components Theory
|
Our recognition of objects is based on features which are the basic units of objects. There are 36 different geons which enables us to mentally represent a large proportion of the objects we recognize.
|
|
|
Principle of componential recovery
|
The ability to identify an object if we can identity it's geons is the principle of componential recovery?
|
|
|
Scene
|
A view of real-world environment that contains (1) background elements and (2) multiple objects that are organized in a meaningful way relative to each other and the background.
|
