Subscribe to Computing Intelligence

Tuesday, October 21, 2008

Project Background Part I: Mirror Neurons

As I mentioned at the beginning of the semester, I am taking a fourth year project course this year. It is a fairly open-ended project, especially since the focus is outside the usual realm of endeavours that my supervisor performs with his research group. However, my supervisor is a great professor in one of the robotic research groups at UTIAS with a great variety of intellectual interests, and hence he was excited about the idea of letting me have a go at coming up with something interesting. It is slightly daunting, since it means a lot of what I am doing I am just kind of muddling through without direction from above, but at the same time it is exciting because it is my question and my problem to search for an answer to.

Anyway, I will get to the actual project in a little while. Rather than have a giant single post which will take ages to write and then likely not be read due to its length, I decided to break it up into some useful background posts first, and then a post on how this background ties into the experimental question I am pursuing. Suffice for now to say that it is centered around developing communication among multiple independent robots (all in simulation, of course, since physical robotic experiments are much more difficult, expensive, and time consuming to perform).

The first part of necessary background knowledge for the project is a thing called mirror neurons. These are a special class of neurons first identified in macaque monkeys by a scientist named Dr. Giacomo Rizzolatti at the University of Parma. These neurons are special in that they fire (neuroscience shorthand for meaning they change their firing pattern and increase firing frequency) in response not only to an individual performing a specific action, but also for an individual watching another perform that same action. For example, when you watch a game of soccer on television and see the slow motion replay of a brilliantly executed shot on goal, somewhere in your head a set of neurons are playing out the same pattern that they would were you on a field running up to a ball and letting fly your best kick.

The existence of such a system of neurons clearly has important implications for both motor learning as well as social interaction. It helps explain such simple psychological phenomena as why people often have the urge to smile when they see others smile, or feel a rush of testosterone filled manliness when watching a mindless action movie (perhaps that's not an event familiar to everyone, but I know it at least works on a childhood friend of mine who, if given the choice, only watches mindless action films. I don't know if he reads this blog or not, but if he does, he knows who he is). It also has many important implications on learning and social behaviour, although those implications will be discussed in more detail when I talk about the project itself. In the meantime, stay tuned for Part II, which will discuss some different control theory models and their applicability to neuronal systems.


Anonymous said...

I had grandmother down at the computer so I showed her your blog. We were reading the blog on neurons and are totally confused. But, I wonder who the friend is? Gadicke, probably.