Object recognition - week 2 (Chris) Flashcards
(26 cards)
Modularity of the visual system: 2 visual pathways
Ventral (what - vision for perception) - parietal lobe
Dorsal (where - vision for action) - occipital/temporal lobe
Milner & Goodale (1992)
What vs where in monkey visual cortex (Pohl, 1973)
Task A - specific pairs of objects predict food reward.
Lesions to inferotemporal (ventral) cortex impair object recognition (‘what’)
Task B - proximity of the cylinder to the food well predicts reward.
Lesions to parietal cortex (dorsal) impair spatial recognition (‘where’).
Occipitotemporal cortex lesions (ventral)
Visual agnosia
Deficit in recognizing objects
Parietal cortex lesions (dorsal)
Deficits of spatial awareness
– hemispatial neglect
Kohler et al. (1995)
In task 1, subjects were presented with two displays and had to judge whether the locations of the objects were the same in both displays.
In task 2, subjects were again presented with two displays and this time had to judge whether all the pictures were the same objects in the two displays.
They found that contrasting the two tasks produced different patterns of activation
Activation was greater for the object task than the spatial task in ventral temporal cortex – primarily fusiform gyrus
Activation was greater for the spatial task than the object task in dorsal cortex, primarily inferior parietal cortex.
Karnath et al (2009)
This experiment shows that the ‘what vs where’ distinction is not quite that simple.
They tested a patients (JS) with a circumscribed lesion to ventral occipitotemporal cortex on two tasks, one requiring a perceptual judgement (task 1) and the other requiring a motor action (task 2).
Performance was compared against non lesion controls
They found that the patient was impaired on the perception task but performed normally on the motor task.
This shows that ventral lesions impair vision for perception but not vision for action, suggesting the ventral/dorsal distinction may be more along these lines.
Note that patient DF was from a previous study (showed similar pattern but had a less circumscribed lesion).
Visual form agnosia
Impairment in visual perception but above the level of a basic sensory deficit (e.g. visual field defect)
Patient cannot recognize, copy, match, or discriminate simple visual stimuli and cannot recognize even simple shapes such as triangles or circles
Associative agnosia
Basic perception seems fine but recognition cannot take place
“A normal percept stripped of its meaning” (Teuber, 1968)
Patient can make good copies of objects but cannot recognize them
Landis et al., 1982
Patient with visual form agnosia read this repeatedly as ”7415” (This)
Inability to group and integrate objects into a whole
Deficit in shape processing
Copying of objects by a patient with associative agnosia
This patient could copy objects very well but couldn’t identify the objects
Not due to a language deficit e.g. anomia
Patient with anomia would be able to describe a picture of a dog, say it’s a pet and that it barks
Patient with associative agnosia would not even know if the dog was an animate or inanimate object
Matching by functions task
Patients required to match the two objects most closely related by function
In associative agnosia, patients are unable to associate items with their functions.
In the matching by functions task, patients are asked to match the two items that are most closely related by function.
Patients with associative agnosia will choose the two most visually similar items indicating that they are unable to retrieve the functions associated with the objects.
Open umbrella, closed umbrella and walking stick.
What does agnosia tell us about object recognition?
Object recognition is modular
Separate brain systems for different processes
Different types of agnosia broadly highlight distinction between:
Perceptual processing (shape analysis) – impaired in visual form agnosia
Semantic processing (activation of associated knowledge, e.g function) – impaired in associative agnosia.
Object recognition is a constructive process
The brain constructs representations of objects based on many different sources of contextual information. These representations, not simply retinal input, are what we are consciously aware of.
Object recognition is a semantic process
Information about the meaning of an object is automatically processed when we see it, e.g. its function
What can agnosia not tell us about object recognition?
Specifically, where in the ventral visual cortex all of this happens
Lesions can be large, diffuse and extremely variable, often affecting multiple regions, even encompassing different lobes
Quite rare to find very ‘pure’ neuropsychological cases
The primary visual cortex is retinotopic
Structured retinotopically
Different regions of the visual field (real world) are perceived in different regions in the primary visual cortex
There is a correspondence (a mapping) between the spatial structure of the primary visual cortex and the spatial structure of the real world
Cortical magnification
A disproportionately large area of the visual cortex is dedicated to the centre of the visual field (corresponding to the eye’s fovea).
V1
Neurons sensitive to simple visual features e.g. line orientation, spatial frequency, colour
LO
Begin to see sensitivity to more complex features such as geometric shape
V5
Plays a key role in motion processing
Kourtzi & Kanwisher (2001)
fMRI evidence has revealed a specialised brain region for integrating features into shapes.
The lateral occipital complex.
In this study, subjects viewed three different types of object – familiar objects, novel objects and non-objects. Only two of these types of objects required the integration of features into shapes. The other objects were just collections of disjointed features.
Activation in the lateral occipital complex, part of the ventral processing stream, was higher for the familiar and novel objects than the scrambled non-objects.
This region plays a role in integrating features into whole shapes.
Properties of area LOC
The LOC is a largely non-retinotopic area, activated by both the contralateral and ipsilateral visual fields
LOC is not simply sensitive to retinal input
Seems to encode higher-level representations of shape, even when not defined purely by retinal input…
Mendola et al (1999)
fMRI evidence that LOC responds to shapes defined by illusory contours (Kanizsa figures)
Researchers have used fMRI adaptation to map different component processes of object recognition onto different regions within the ventral visual pathway…
Can use this to investigate sensitivity of groups of neurons to different object properties.
e.g. this neuron is sensitive to object identity – it ‘knows’ the pictures are all of the same object as it reduces its response with repeated presentations
The fact that this reduction in response occurs even when the repetitions occur from different viewpoints means the neuron is insensitive to viewpoint
Vuilleumier et al. (2002)
fMRI reveals a region in left fusiform cortex that represents objects in a viewpoint independent manner – this region is sensitive to object identity, regardless of viewpoint.
Looked at fMRI adaptation – the extent to which activation decreases with repetitions (when you present the same object or word twice, activation tends to decrease – neurons ‘adapt’ their responses to the object – by varying different properties of stimuli you can assess the extent to which a brain region processes that property). A simple example would be tones of different frequency. You might present tones of different length and different frequency. Neurons that process frequency will adapt their response to tones of the same frequency even if the length of the tone differs.
Found reduced activation in left fusiform cortex to the same object from a different viewpoint relative to when different objects were presented.
Thus, this region treats the top two pictures the same even though they’re presented from a different angle, indicating this region may play a role in object constancy – enabling us to recognise an object under multiple different contexts, viewpoints etc.
