Schedule
| Wed 5,6 (2:50-5:50pm) | Psych A139 (Busch) [map] |
![[map]](img/toplayroom.gif){kind=link}
Staff
| Instructors: |
Prof. Doug DeCarlo Prof. Matthew Stone |
| Email: | decarlo@cs mdstone@cs |
|
Schedule
Staff
|
|
Links
Over more than fifty years of study, researchers in artificial intelligence and cognitive psychology have explored a range of processes that creatures must have to act intelligently in the world: intelligent creatures have to be able to understand their environment; they have to be able to make plans and draw conclusions; they have to be able to learn from experience; and they have to be able to put new ideas together creatively. There is no single key to general intelligence - which is why the sophisticated robots of science fiction are still unlikely to materialize any time soon. In fact, when intelligent computer systems finally arrive, they will probably include a diverse array of special-purpose mechanisms that make them precisely sensitive to a wide range of complex relationships in the world.
The goal of this seminar is to introduce the practical grounds for this view of intelligence through a series of engaging visual case studies. We will draw on the formal study of human perception, action and communication to create and study computer programs that, to some degree, can understand and create artistic images. This is a self-contained class which requires no prior programming experience. Unlike typical computer science classes, we won't emphasize the brute mechanics of getting computers to do things. Rather we will emphasize representation: the practice of describing situations in the world in a precise way that computer systems can then use. A precise description is actually enough to allow a computer to simulate something or reproduce it. But precision is key, because a computer cannot give you any slack.
Our case studies will include embodied utterances, which pair words with appropriate intonation and facial expressions - like people use in face-to-face conversation. We'll read Faigin's An Artist's Complete Guide to Facial Expression, an artistic perspective on the ways the face conveys meaning, emotion and personality. And we'll implement descriptions using our research prototype RUTH (the Rutgers University Talking Head). We'll look at sequential art (aka comics); we'll read McCloud's Understanding Comics, a comic book about the meaning of comic books, and we'll see what a computer system has to represent to string together images so they exhibit a recognizable chain of cause and effect. Finally, we'll look at the structure of visual designs, reading Holtzmann's Digital Mantras, a survey of computers in language, music and art, and see how computers can create abstract images with an accessible visual structure using recursive rules.
Take a peek at the cereal lab information now available. You should have received email from us about your cereal account -- let us know if you didn't.
Here's a link to the DrScheme environment that we used to demonstrate the Scheme language in class Wednesday. It's free and easy to install on your personal machine if you'd like to play with it. For example, you can use it to type in the code from the book, The Little Schemer.
We hope that accounts in the instructional lab (which has DrScheme and other software) will be set up by Monday.
If you don't have the book yet, just click on the link below (it is a PDF file; you'll need the Adobe Acrobat Reader to view or print a PDF file. All computers in campus labs will have it. Get it here if you need it for your personal machine.)
Also, be sure to take a look at this map so you can easily find the classroom!
DrScheme examples:
Go through this example; it should help you understand maximum from class. You can load it in "beginner scheme with list abbreviations" and step through it.
DrScheme examples:
Links: NYU Face
DrScheme examples:
DrScheme examples: