Week 5 Object & Face Perception

Question 1

What are the differences in receptive fields of M and P cells?

Answer 1

M cells have large receptive fields of a visual scene (than P-cells) to give you course detail of stimuli

P-cells have small receptive fields of a visual scene, allowing them to be highly sensitive to high spatial frequencies and finer visual acuity

Question 2

What are the differences in temporal frequency fields of M and P cells?

Answer 2

M-cells are highly sensitive to rapid changes in temporal luminance (high temporal resolution, ie. fast movement and flickering

P-cells are less sensitive to temporal changes (movement, motion and luminance changes over time)

Question 3

Which cells, M or P, are colour sensitive?

Answer 3

P-cells are colour sensitive, M-cells are colour blind/monochromatic

Question 4

Where do M and P cells go and which is faster?

Answer 4

M-cells go to the dorsal pathway and have a faster transition speed than
p-cells, p-cells go to the ventral pathway

Question 5

What does Mishkin & Ungerleider 1982 say about the dorsal and ventral pathways?

Answer 5

Dorsal pathway = object location (the ‘where’ pathway)

Ventral pathway = object perception (the ‘what’ pathway)

Question 6

What are 3 brain regions that encode objects in the ventral stream?

Answer 6

1. Fusiform gyrus (sensitive to faces)
2. Parahippocampal gyrus (sensitive to places)
3. Auditory, visual and visual word form areas in the ventral stream

Question 7

What are 2 brain regions that encode object movement in the dorsal stream?

Answer 7

1. Parietal cortex = spatial attention, encoding objects in scenes, where and when to pay attention
2. Motor cortex = visually guided reaching, ie. holding, grasping, figuring out the size, shape and density of objects

Question 8

What is the “frame and fill” approach for encoding a visual scene?

Answer 8

1. Visual information from environment gets encoded in a feedforward sweep from the retinal image to the V1 to other posterior occipital regions to anterior temporal (object perception) and frontal regions
2. This feedforward sweep is generating a perceptual hypothesis about what the object is
3. To confirm the perceptual hypothesis, information is encoded in a feedback/reentrant process, whereby visual information in frontal areas goes back to occipital areas like V1 to compare high resolution information with the initial hypothesis
4. If the perceptual hypothesis is deemed correct, this is what is consciously perceived

Question 9

How does Bar (2003) apply the frame and fill approach for M and P cells?

Answer 9

1.. Eye sees a visual stimulus, a magnocellular shortcut gets projects low-spatial frequencies directly into the prefrontal cortex
2. M-cells have low spatial frequencies, so the visual representation will look fuzzy/course
3. This triggers a feedback mechanism back to the LGN (that receives more feedback than feedforward) to the ventral stream and the temporal cortex to clarify the object with input from P-cells

Question 10

What 4 pieces of evidence supports Bar’s model of visual processing?

Answer 10

1. Faster responses
   Activity in the orbitofrontal cortex - where M-shortcut is believed to occur was processed quicker (50ms)
2. Predicted object recognition
   in orbitofrontal cortex compared to more classical object-related areas like the fusiform gyrus
3. The orbitofrontal cortex had higher activation for low-spatial frequency images, not high-frequency images
4. Greater synchronisation between feedforward and feedback pathways for low spatial frequency images compared to high spatial frequency images

Question 11

What are the similarities/differences between Bar’s model and Navon’s global precedence idea?

Answer 11

Similarities:
Global letter = processed by the magnocellular cells the overall global shape
Local letter = processed by the parvocellular cells

Explains why we are faster at processing global letters regardless of the letters are congruent/incongruent

Differences
Bar’s model makes the distinction between object distinction and feedforward and prefrontal cortex feedback processing, while Navon’s idea focuses on the visual outcome of what is processed first

Question 12

What is facial recognition vs. facial identification?

Answer 12

Facial recognition = is it a face?
Facial identification = whose face is it?

Question 13

What are the differences in arguments for and against face processing?

Answer 13

Faces are unique -
Unique brain area exclusively dedicated to facial perception and not other objects
Inversion effects occur because we have holistic processing of faces, we do not process them in isolation
Global processing is unique to faces

Faces are not unique -
Faces are one of many stimuli that use holistic processing
So-called face-specific processing areas are actually sensitive to recognition things you already know

Question 14

Are inversion effects for faces good evidence that faces are unique?

Answer 14

No - while inversion effects in faces seem to have a preferential effect for holistic processing, it is not indicative that only faces are only processed globally

Other stimuli that people have expertise in, such as words, houses, and made-up stimuli (greebles) also show inversion effects

Question 15

Why is the fusiform gyrus area (FFA) thought to be responsible for face processing?

Answer 15

fMRIs indicate that when faces were presented, higher activation in FFA compared to other objects, scrambled faces, hands and houses

Most critically, the lower activation of the FFA with scrambled faces indicates that the FFA does not simply respond to individual facial features, but rather the holistic arrangement of a face

Question 16

How could a neurological case support the idea that FFA are only involved in faces?

Answer 16

1. Find cases of double dissociation where a brain injury to the FFA leads to an impairment of facial processing but not other stimuli or expertise stimuli

Question 17

What is the limitation of the Hemke (1998) study on prosopagnosia?

Answer 17

Limitation of this study is that it is easier to visual discriminate between categories (chairs vs tables) than within categories (faces vs faces) - ie. degree of difficulty is not the same between conditions

Question 18

What dog study supports the idea that the inversion effect is for objects of expertise and not just faces?

Supports the expertise hypothesis/faces are not unique

Answer 18

Diamond & Carey (1986)
The dog-experts showed a large inversion effect with inverted dog images, but novices did not show an inverted effect for dog images

This supports the expertise hypothesis, that dogs as well as faces can also be susceptible an inversion effect, if an individual has a high expertise in dogs

Question 19

Why is the inversion effect thought to occur?

Answer 19

The inversion effect occurs in highly familiar stimuli, such as faces or dogs, because it greatly impairs the processing of holistic information.

Question 20

What is the FFA most likely involved in?

Answer 20

Processing complex stimuli that are hard to discriminate from one another

by processing the spatial relationships between features

Question 21

How to test the expertise hypothesis of the FFA? (Greebles)

Answer 21

1. Gather 2 groups of participants - make one group practice discriminating with complex stimuli, leave the other group as novices
2. Measure both groups FFA activation in an fMRI machine when presented with the complex stimuli and compare FFA activation between novices and experts
3. Finally, compare FFA activation in experts between their learned object and other control stimuli

Question 22

What were the two main results of the Greeble study?

Answer 22

1. Greeble novices did not show much activation of the FFA when presented with greebles
2. The greeble-experts showed similar FFA activation in greeble stimuli as with face stimuli.

Rossion et al. (2002) found N170, ERP component for facial recognition, was elicited in greeble experts

Question 23

How can people with prosopagnosia use other ways to recognise someone?

Answer 23

1. Try to remember body shapes, styles of clothing, name tags.
2. Since the ventral pathway is modality specific, ie. people with visual agnosia would be able to recognise objects by using other senses, ie. touch, smell, hearing.

Question 24

What could super recognisers use their ability for?

Answer 24

Good for security, airports etc. ie remember who has come in or out already, and matching faces to fraudulent passports, drawing people, missing people’s cases.