For your final course project, you should design and implement a research prototype that:

• Uses data-driven methods for a computer graphics problem,
• uses methods from computer graphics for a machine learning problem, or
• combines computer graphics, computer vision, and machine learning in some other novel way.

The project should consist of a novel research prototype or experimentation. It should be thought of as a proof-of-concept: something that, ideally, could be continued after the semester and then published in a conference or journal. (Of course, you are under no obligation to do so).

You may use whatever resources you want, such as external source code, data, or conversations with other people. I've placed a number of resources on the course web page, including data sets and numerical libraries.

Project ideas: You are encouraged to devise an interesting project idea. Be creative! I will also provide a list of ideas in class that you might also find helpful. I will also be happy to propose ideas if you have an idea of what area you want to work in, but are not sure what to do. You might also find papers on the web or datasets that provoke interesting ideas.

Project meetings: You should discuss your project idea with me by November 3. However, I would encourage you to begin thinking about the project well before this date. If you wish, I will also be happy to discuss your project as it continues through the semester, time permitting. I will be out of town on September 14-25.

Groups: You are encouraged to work in groups of two, but it is also perfectly acceptable to work alone. Talk to me if you prefer a larger group.

Final presentations: The projects will be presented to the class in December (tentatively: December 6-10). Let me know if you have scheduling constraints. Your presentation may include PowerPoint, live demos, or whatever else best describes your work. You do not need to submit any written report.