http://italkproject.org/
This video demonstrates how Aquila trains and runs multiple time-scales recurrent neural networks used to control humanoid robot iCub (http://www.robotcub.org/) in real time.
The results shown in this video are from the initial tests only and are meant to demonstrate the potential of neural system (based on Jun Tani's architecure) as well as the massive speed-ups achieved by running the network on parallel CUDA devices such as the latest NVidia Fermi architecture.
We are currently focusing on extending the neural system with additional modalities dealing with vision, extra proprioception as well as language so keep eyes on my site for new results. I reckon it is going to get more interesting as we proceed farther with our investigation into how actions and language co-develop in the brain.
@wlorenz65 Think about as a child learns stuffs. First of all, you need to recognize that is your hand, learn the limits of it, then how to use it. It's a really long procedure, like as we learnt in our childhood. That alternate when it will study by seeing the manouver is a bit further, but the oldest one has got a new skin nowadays, as it will learn the forces, and how strong or smooth he may grab something, (not to break it) it will be more efefctive. But gosh its so cute :D
saduntrist 4 months ago
creepy
jorgento 1 year ago
What happens if the blue box is replaced by one twice the size which has a flower pattern on it? Will iCub be able to replay the learned actions?
I would really like to see a video where a HUMAN operator is doing the stuff and iCub is watching and then iCub will replay the actions. Well, a long time until then ...
wlorenz65 1 year ago