CAS PS 222 Lecture Notes - Lecture 7: Receptive Field, Retinotopy, Lightdark
18 May 2018
School
Department
Course
Professor
Lecture 7: V1
If we want to see objects in the world that are behaviorally relevant- like paintings, if we need
to see the world, we need to first identify the physical stimulus that is light in this case.
How does light carry information? - different wavelengths of light helps us distinct items In the
scene. They carry spatial information. The Spatial information is the reflectance properties of
the detailed aspect of an image or object in the world. Two distinct objects that behave
separately and are on their own.
The location of the differences in the reflectance- the hand is here and the face is there- this
helps us deal with the spatial organization of light in the world. We see the world unscrambled.
The reflectance properties are important in dictating where the objects and the items are.
We have to now funnel the physical stimulus to the site of transduction- photoreceptors.
There is an upside down left-right reserved 2D image.
As we move away from the physical stimulus and towards the neural signals.
Somethings can be too bright for the photoreceptor to detect. But all this gives us is little
iforatio aout the lightess/darkess i the iage. We dot reall ko the iage o.
The photoreceptors are the first stop. They then send the signals to RGC.
‘GC dot hae the sae ork as photoreeptors. Ho does the ‘CG respod to stiuli i the
surrounding? – they detect light depending on where it is in their receptive field: center-
surround.
RGC receives signal from photoreceptors about the pixels and then start to organize them in
the brain. They organize the light/dark boundary. The pixels are organized into light/dark
boundary.
RGC receptive fields size vary
Smallest RF: Fovea
Largest RF: Periphery
There is a distribution in the RGC RF which gives us high resolution in the fovea and opposite.
Where do the axons of RGC project to? – Their axons form the optic nerve and the first stop is
LGN.
What is the functional role of LGN? – LGN integrates information from both eyes. We have 2
photographs of the world taken from two different angles so each eye is carrying slightly
redundant information but with a little offset which is going to be important.
The pictures from two eyes are over-layered. Retinotopic organization is maintained in the LGN.
It further increases the detail of boundaries as compared to RGN.
It is important to integrate the images of the 2 eyes because each eye has a 2D image of a 3D
world. So how do we change the 2D image to 3D? – So, in LGN retinotopic layering is
maintained. We are not even the cortex and there is so much organization already. But we still
need to build shapes from this information
LGN is in the midbrain in the thalamus.
find more resources at oneclass.com
find more resources at oneclass.com
As we move to the cortex- we do some heavy-duty organization.
V1 is where a significant processing is done where shapes and forms, depth, motion starts
taking place.
What is the functional role of V1 is visual processing?
- To understand it we need to understand these two questions
- What is the preferred stimulus of V1 cells
- What is the organization of V1?
What are the preferred stimulus?
- We need to use the process of sticking an electrode in the V1 cell of a monkey, and
recording from the cell when light (stimulation) is shined on it. Measure the cells
response as you show stimulus to the animal. A single V1 cells reaction to a bar of light
is shown in the slides.
- As you change the orientation of the bar, there is difference in activity of the V1 cell.
- The preferred stimulus is a vertically oriented bar. How do we know? – The activity goes
up the most when a vertically oriented bar is in the receptive field of the cell.
- You can map out the cells average response to stimulation. X axis could have orientation
of the bar. And on Y axis the response of the cell. You end up with a distribution of the
stiulatio ith its peak at the ertial ar. So its a ell shaped graph.
- How do you think the cell would respond if the bar is horizontal? – The activity of the
cell remains the same.
- There may be many different cells which have different preference to the orientation of
the bar.
Why is orientation important? Why is it important to be able to detect a variety of different
orientations? How are we able to detect shape boundaries then?
- V1 gets its information from the boundary detectors in LGN so it responds to
orientation. We can map out the receptive field of a V1 cell, and if we map it out and we
find a receptive field that makes the cell sensitive to orientation.
In the Simple cell:
- Very well defined on/off regions
- If the stiulus falls o the + regio, the aitiit of the ell ireases
- If the stiulus falls o the - regio, the atiit of the ell dereases
- If it is uniformly illuminated, baseline firing occurs.
How do we go from center-surround organization to orientation detection?
- All cells respond on center-surround organization for most parts. How do we change
that to orientation detection? V1 gets its signal from LGN which is not just a one-to-one
mapping between the cells. V1 could be getting variety of sources of LGN. Multiple LGN
could be sending information to one V1 cell.
- The LGNs have their receptive fields on a locally organized site of the retina. These LGN
cells have receptive fields that are adjacent or on each other in the retina.
- If we were taking our inputs from multiple LGN cells which has its center surround RF
we can build this orientation detection because V1 can use all the information from the
multiple LGN signals.
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
If we want to see objects in the world that are behaviorally relevant- like paintings, if we need to see the world, we need to first identify the physical stimulus that is light in this case. Different wavelengths of light helps us distinct items in the scene. The spatial information is the reflectance properties of the detailed aspect of an image or object in the world. Two distinct objects that behave separately and are on their own. The location of the differences in the reflectance- the hand is here and the face is there- this helps us deal with the spatial organization of light in the world. The reflectance properties are important in dictating where the objects and the items are. We have to now funnel the physical stimulus to the site of transduction- photoreceptors. There is an upside down left-right reserved 2d image. As we move away from the physical stimulus and towards the neural signals.
