![]()
Visiting guest. Why not sign in?
|
|
Reply to Message
Not registered yet?
The AI Depot has a focused community of friendly users. Rather than let anyone abuse the site at the brink of promiscuity, we prefer to let only those with an active interest participate... this simply requires registering.
Why not sign up!
Joining the site's community is completely free. You can then post messages freely, and customise your personal profile at will. Specific privileges will also be granted to you, like being able to access printer-friendly articles without restrictions. So, why not register?
Parent Message
Jackson Pollocks?
As far as I'm aware art is mostly about highly abstract ideas, and I don't think there are any robots around at present capable of having such ideas. The paintings may be of interest to the people, but I'm sure they're of no consequence at all to the robots themselves. The ability to value artistic creations seems to be uniquely human. I know some people have trained chimps and even elephants to paint, but the animals themselves don't ascribe any value to their creations. They don't exchange them with eachother as a valued comodity like humans do. - Bob |
|
Couldn't say it better myself...
Bob, I can't agree more. The robots described have no understanding of what they are creating. My PhD thesis "Curious Design Agents and Artificial Creativity" makes this very point with respect to agents in art and design -- available at http://www.arch.usyd.edu.au/~rob -- i.e. for artificial agents to create art or design creatively they must have some understanding of what they are doing. The eponimous curious design agents are my initial attempt to make the novelty of works important for the agents. The development of Artificial Creativity systems (along the same lines as Artificial Life systems) mirrors your comment that artists exchange ideas and works as valued commodities. The most striking thing about the paintings done by elephants is that the paintings by a single elephant often display a definite style. I think this is interesting not because I believe that the elephants are trying to express themselves but because it may indicate that they have a preference for certain patterns or painting actions. Then again it could be a reflection of their individual training and their trainers preferences. Who knows? BTW: Jackson Pollock was attempting to replicate the patterns he saw in nature without producing representations of natural objects. In essence he was searching for a way to capture the fractal dimension of the world around us. Research measuring the fractal dimensions of Pollocks' paintings over time showed that as he perfected his skills he came closer to approximating the fractal dimenion of natural scenes. -RoB- |
|
ART for art's sake
That's an interesting paper and some nice applets. I'm interested in the ART algorithm because it might be useful for my robot's vision system. At the moment the robot has a short term visual memory which is maybe similar to the simple ART algorithm, except that the vectors don't move at all and the tollerance is a global fixed parameter. It might be interesting to try using something like the classification ART algorithm so that the tollerance values may vary from one classification to the next. - Bob |
|
Category ART
Feel free to have a look through the source code provided on the applet pages. I'm afraid that I never really got around to commenting them better but the algorithms are quite simple. David Weenik's papers on his modified ART algorithms easy to follow if you'd rather code from the original source. In addition, you might consider using Growing Neural Gas algorithms that are also mentioned on the applet pages. -RoB- |
|
Classifying visual inputs
I had a look through your code and I've modified Rodney's STVM accordingly. I do remember seeing a video a long long time ago where (I think) Grossberg described his ART algorithm. It was a long boring highly mathematical lecture and I never payed much attention at the time. But reading your paper it seemed like the classification version of ART would be a neat solution to classifying the robots visual input. The expanding and contracting tollerances (receptive fields) mean that the robot can produce fine classifications for objects which are more interesting or more frequently observed, and fairly broad classifications over areas of the feature space which aren't very interesting (such as Grossberg lectures). I don't know exactly how you did it but I just wrote an algorithm such that each "neuron" tries maintain a constant number of neighbours within its receptive field. If there are not enough neighbours the field expands. If there are too many neighbours the field contracts. - Bob |
|
Good idea
Bob, Trying to maintain a constant number of neighbours per neuron sounds very reasonable. One of my research aims was to show that a model of curiosity could be built with standard components so I just implemented the algorithm as published -- reducing the receptive field by a fixed percentage (eg. 10%) as needed. I'd be interested to know how you get on with your variation. -RoB- |
|
Back to the Artificial Intelligence Depot.