Project 1–Phase 2

by Mike Gleicher on March 2, 2011 · 1 comment

in Assignments,Project 1

Now that you’ve built a bunch of infrastructure and read a bunch of papers, its time to put the two together and do something interesting!

Here is the rough plan:

  • Friday 3/4 – Each group puts together a plan. A posting to the “Project 1 – Phase 2” category describing what you intend to do.
  • Friday 3/11 – Each group gives an update: updating the plan, and showing their progress (there had better be some)
  • Friday 3/18, Friday 3/25, Friday 4/1 – there are 4 weeks from the Checkpoint to the Final. I expect everyone to take some time off (either break, or to work on a deadline, or …). I will expect updates along the way – details to be provided.
  • Friday 4/8 – Project Due. We will have demos, as well as written project descriptions. Details coming soon.

It seems like most groups have decided to continue to work together closely. Remember in the initial assignment, I had encouraged groups to split into 2 parts (either 2 and 2, or 1 and 3), but to help each other supporting infrastructure. This is your decision.

Types of Projects (these are not disjoint – a project might fall into several categories)

  • Recreate an existing technique in order to gain a deeper understanding of it. Note: your result here should be a way of helping others learn as well, such as a nice tutorial or explanation.
  • Build a special-case version of some complicated thing, focused on understanding it or trying it in a new setting.
  • See how two techniques fit together (motion graphs + path editing, motion graphs + splicing)
  • Implement multiple techniques and compare them
  • Explore the use of different representations with standard techniques (e.g. blending on point clouds)
  • Try an existing method with some different design choices (fragment-based assembly with varying time steps, motion graphs with a different matching or transition generation, …)
  • Invent a new method for an existing problem
  • Create a practical pipeline that feeds a simple version of a method (for example, do topology conversion / retargeting so you can use varieties of motions with motion graphs)
  • Revisit one of the basic challenges (transitions, retargeting, …) and explore the space of known solutions – and maybe find a new one.

I mention those to be clear that there are many different ways for a project to be interesting.

If you have questions about what is appropriate, talk to me.

Previous post:

Next post: