Lecture 14 - Object Recognition: From Objects to Faces Flashcards
What is one of the main benefits of having two eyes?
Two eyes means that we can have a perception of depth.
What is Marr’s theory of Recognition?
Marr’s theory of Recognition looks at the brain as behaving like a computer (computational model), developing algorythms that allow for recognition.
Marr’s theory of Recognition says that one of the first stages of recognition is the recognition of EDGES.
This recognition of edges occurs at the level of the retina and the LGN.
‘Edge’ detectors are receptive fields that are sensitive to light discontinuity - characteristic of edges.
True or false?
True.
At the level of the Primary Visual Cortex (V1) there are orientation receptive fields.
True or false?
True.
What occurs at the level of V4 in the visual cortex?
Neurons in the V4 are receptive to curvature and shape in the light signals being received by and transferred from the retina.
What does “viewer-centred” representation mean?
Viewer-centred
What is the role of the Inferior Temporal Cortex in object recognition?
Development of a robust perception of the light signals that are
Recognition of objects requires prior experience/learning and memory.
This means that object recognition requires not only the visual cortex, but connections to other areas of the brain.
True?
True.
Is processing of light signals in the Inferior Temporal Cortex viewer-dependent?
No. The processing and reconstruction of light signals at the level of the Inferior Temporal Cortex is a more robust representation of what we are actually “seeing” and is less affected by movement or changes in direction of the visual field - this means we can still recognise something even if we or it moves.
This is not the case for the processing of signals that occurs at V1 and V4.
What is the Ventral Stream?
In regards to visual processing the ventral stream stream is known as the “What” pathway for object recognition.
There is evidence that suggest modularity of the visual cortex when it comes to areas that are more involved in the recognition of certain types of objects, such as the Fusiform Face Area.
True or False.
True.
What is visual agnosia?
Visual Agnosia is the inability to recognise objects. This is often seen in individuals who have lesions or damage to the Inferior Temporal Cortex. The extent of a person’s visual agnosia in these cases in dependent on the size of the lesion, which lends evidence to the idea of modularity of this brain area in regards to object recognition.
Humans are very good at recognising faces compared to other objects.
True or false?
True.
There are neurons in the Inferior Temporal Cortex that are more active in response to faces and configurations that look like faces.
True or false?
True.
What is prospagnosia?
Prosopagnosia is the inability to recognise faces.
The brain is also able to recognise contexts and scenes, aka scene recognition. What area is responsible for scene recognition?
The Parahippocampal Place Area.
What is the “inversion” effect?
The inversion effect refers to our inability to recognise faces and scenes as well if they are upside down/inverted. as we normally see faces or scenes upright.
The brain is able to store a large amount of visual information from scenes with high fidelity even if the exposure to this scene is brief and there are many other competing scenes.
True or false?
True.
Is object recognition a precursor for scene recognition?
No. Scene recognition can still be in tact even if there is damage to the Inferior Temporal Cortex (Area for object recognition).
Where in the brain does evidence suggest scene recognition occurs?
Scene recognition is postulated to occur in the Parahippocampal Place Area (PPA).
What is a hemianopic field?
Hemianopia is the loss of vision in half of the visual field in one or both eyes. Most commonly caused by a stroke.
What is blindsight?
Blindsight refers to the unconscious processing of visual information, such as knowing an object moved, but not “seeing” it move, if you have a certain cortical blindness.
What is the first thing that is required in object recognition, according to Marr?
EDGE recognition. Edge-sensitive receptive fields are found initially in the LGN.
Orientation coding at the level of V1 allows for the detection of edges and their continuity or discontinuity, according to Marr.
True?
True. At V1 we are processing information not just information about whether there was an edge, but also what orientation the edge has.