Object Perception Flashcards
What is Agnosia?
failure to recognize objects even though you can see them
What were the characteristics of cells in IT (inferotemporal cortex) discovered in monkeys?
They were discovered to have receptive fields that could spread over half the monkey’s field of view.
What are the two main pathways for visual information from the occipital lobe?
“where” pathway and the “what” pathway.
What does the “where” pathway process and where does it go?
The “where” pathway processes information related to the location of objects in space and actions required to interact with them, and it goes up to the parietal lobe.
What does the “what” pathway process and where does it go?
The “what” pathway is responsible for explicit acts of object recognition, and it goes down to the temporal lobe.
What is the role of the “where” pathway in attention?
The “where” pathway plays a role in the deployment of attention.
What happens to receptive fields as we move into the temporal lobe?
Receptive fields become larger as we move into the temporal lobe.
What is the importance of the identity of the object in the “what” pathway?
`
In the “what” pathway, the identity of the object becomes more important than its location.
What parts of the brain is the IT cortex closely connected to?
The IT cortex is closely connected to parts of the brain involved in memory formation, such as the hippocampus.
How do respective fields and simple features of the visual cortex change as one moves from the striate cortex to the IT cortex?
They become more complex as one moves from the striate cortex to the IT cortex.
What is a “grandmother cell”?
Any cell that seems to be selectively responsive to one specific object
What is the Extrastriate Body Area (EBA) responsible for?
Responsible for being activated by body structures other than the face.
What is the Parahippocampal Place Area (PPA) responsible for?
Responsible for cells that respond to places in the world like rooms with furnitue
What is the Visual Word Form Area (VWFA) responsible for?
The Visual Word Form Area (VWFA) is responsible for recognizing visual written words.
What is a feedforward process?
A process that carries out a computation, such as object recognition, one neural step after another, without the need for feedback from a later to an earlier stage.
What is the reverse hierarchy theory?
The reverse hierarchy theory suggests that instead of processing information from small details to big picture, our brain processes information from the big picture to small details.
What do high-level visual areas do to obtain more details about objects?
High-level visual areas send signals back down to low-level visual areas to obtain more details about objects.
What is high-level vision?
High-level vision is the processing of complex scenes and the integration of information from multiple sources, including specific learned objects or patterns.
What is low-level vision?
Low-level vision is the processing of basic features such as edges, colors, sine waves, and textures in an image.
What is mid-level vision?
Mid-level vision involves the interpretation of basic features to identify meaningful objects or regions in an image, such as contours, group surfaces, figures, and borders. This includes tasks such as object recognition, motion detection, and depth estimation.
What is the structuralist theory of perception?
argues that perceptions are the sum of atoms in sensation, such as bits of color, orientation, and edges
What is the gestalt theory of perception?
The gestalt theory of perception argues that the perceptual whole is more than the sum of its sensory parts.
What are the gestalt grouping rules? (6)
similarity, proximity, good continuation, texture segmentation, parallelism vs. symmetry, and figure-ground assignment.
Gestalt Grouping: Similarity
Tendency of two features to group together will increase as the similarity between them increases
Gestalt Grouping: Proximity
States that items near each other are more likely to group together than are items more widely separated
Gestalt Grouping: Good Continuation
Figures with edges that are smooth are more likely seen as continuous than edges that have abrupt or sharp angles.
Texture segmentation:
Refers to the process of visually separating different areas of an image based on differences in texture. I
Figure-ground assignment
Process of determine that some regions of an image belong to a foreground object (figure) and other regions are a part of the background (ground)
What are two of the strongest principles of texture segmentation?
Similarity and proximity
Parallelism
Parallel contours are likely to belong to the same figure
Symmetry
Symmetrical regions are more likely to be seen as a figure
Reliability:
The degree to which two line segments appear to be a part of the same contour
Global Superiority Effect
The properties of the whole object take precedence over the properties of parts of the object
Bayesian Approach:
Theoretical approach that takes into account the likelihood of perceiving something given your previous experience of the world
When do Y junctions almost always occur?
Y junctions almost always occur when one surface occludes another.
What do Y and arrow junctions almost always correspond to?
Y and arrow junctions almost always correspond to corners and thus don’t signal occlusions.
What are these various junctions an example of?
These various junctions are examples of non-accidental features.
What is border ownership?
Refers to the perception of the location of an object’s border relative to its background. It helps the brain understand the different objects in a scene and make sense of what is being seen.
Why would a border ownership cell in V2 respond more strongly to the edge of a black square than to the locally identical edge of a gray square?
The stronger contrast of the black square makes it easier for the border-ownership cell to detect and identify the edge and respond more strongly.
What are border-ownership cells?
Border-ownership cells are neurons in the visual system that respond selectively to the presence of an edge at a particular location within the visual field and indicate which side of the edge “owns” the border, that is, which side is in the foreground and which is in the background.
Where do border-ownership cells in V2 receive input from?
Border-ownership cells in V2 receive input from simple cells that respond to edges of different orientations and positions within the receptive field.
