— 11 October 2006 —
Okay, not really. But the guy in this video is basically drawing things my daughter would draw, then turning them into complex models.
This is a very cool demo and the fact that MIT is just down the street from me now is even more cool. Maybe I’ll saunter down there and see if I can play with it.
The guy in the video is Randall Davis, PhD (Stanford in artificial intelligence) and is currently the MIT Computer Science and Artificial Intelligence Laboratory research director. There’s a project page associated with this technology that points to a MIT/Microsoft collaboration. For use in their Tablet PC at least from a consumer perspective. They call it “Magic Paper.” Damn marketers… I prefer “Assist Sketch Understanding System.” Seriously. Why not call out across the lab, or conference room, “Hey guys… to demonstrate this better, maybe we should play with our ASUS?” I shoulda gone to business school; I woulda rawked!
Okay, okay. Offically, without marketing’s influence, it’s called ASSIST, A Shrewd Sketch Interpretation and Simulation
Tool. I like ASUS better. Maybe I should rethink that last sentence.
Here’s some extra reading on the topic at hand (PDFs):
- Speech and Sketching for Multimodal Design
- HMM-Based Efficient Sketch Recognition
- Dilemmas Faced by Creative People in IT Only slightly related, but interesting.
I recall using some aspects of this technology back in 1997 on my Apple Newton. There was no complex movement to be generated, but the system could correct my “almost triangle” into a real triangle.