Lecture 7

Question 1

Hierarchical model of representation of objects in the brain

Answer 1

There is feedforward convergence of hierarchical RF properties: from cells detecting low level features to cells detecting increasingly complex feature constellations, to cells detecting highly specific objects (pontifical cells, or grandmother cells).

Question 2

Dynamical assembly formation model of representation of objects in the brain

Answer 2

Cells each encode specific low and higher level features, but objects are encoded by the formation of dynamic assemblies via horizontal and feedback connections, that together encode a particular object

Question 3

Which model of object representation does the visual brain use

Answer 3

It uses both models. There is clear evidence for hierarchical processing, leading to ever more complex RF tuning. There is also clear evidence for dynamic grouping via horizontal and feedback connections to bring it all together

Question 4

When is there hierarchical processing and when dynamical

Answer 4

For very specific objects like faces, the brain adopts hierarchical processing. For most other objects it is dynamical.

Question 5

Combinatorial explosion

Answer 5

There cannot be a cell for every possible object

Question 6

Ensemble coding theory

Answer 6

A potential solution to the combinatorial explosion problem.
It states that an object is not represented by a single neuron but by a set of neuron, where each group encoding a particular feature of the object.
This would also explain that brain damage can’t cause you to forget someone if the neuron representing them would be destroyed.

Question 7

Sycnhrony label for assembly coding

Answer 7

A potential solution to the problem of how does the brain label certain assemblies.
This states that when an assembly is activated all the neurons in this assembly roughly fire in synchrony.
This reflects orientation similarity but not figure ground segmentation.

Question 8

Is decoding faces depending on high or low level areas

Answer 8

Low level areas are just as important in face identity recognition as higher areas.

Question 9

Dorsal stream of visual information

Answer 9

Superior longitudinal fasciculus -> Posterioparietal cortex.
Magnocellular dominated.
Motion and position are important in this stream.
Vision for action.

Question 10

Ventral stream of visual information

Answer 10

Inferior longitudinal fasciculus -> inferior temporal cortex.
Parvocellular dominated.
Colour and shape are important in this stream.
Vision for perception.

Question 11

Associative agnosia

Answer 11

Patients cannot name the object that they see but they can recognize what they feel so they still know the name.
They have no problem with copying drawings and their perceptual organization is fine.
Caused by mostly left hemisphere occipital/parietal/temporal lesions.

Question 12

Ebbinhaus illusion

Answer 12

Perception thinks that the circles are different sizes. However visually guided action does not think this

Question 13

Real-time vs Delayed grasping

Answer 13

Immediate grasping does not suffer from the size contrast effect. Delayed grasping does.
This experiment showed that perception and memory guided action suffer from visual illusions but immediate visually guided actions do not.
This is evidence for the vision for perception and vision for action pathways.

Question 14

Explanation for why illusions don’t work for visually guided action

Answer 14

Vision for action system is the connection between eyes and hands, it is used for immediate visually guided action.
Vision for perception system is the connection between eyes and later action, this is for memory.