Guidelines for paper presenters:

Review the basic ideas of the paper: what problem is addressed,
what strategy is taken, what the paper contributes to the field.
Check the web to see if there are supplemental materials
available, like movie files. If so, it's a great idea to include
those in the presentation. 

Other material not included in the original paper can also help a
lot, like when Alex K. used Photoshop to prove that the light
colored square in the shadow really was the same color as the
dark colored squares outside the shadow (in Durand's paper).

You are also free to bring up other papers that the whole class
didn't read, such as papers listed under "Also of interest" on the 
papers page:
  http://www.eecs.umich.edu/courses/eecs598-5/papers/

Last, think about what could make interesting discussion points.
Often, questions from the review form can work well for
this. E.g., on the question of the limitations of a method,
sometimes just one or several students will spot an important
limitation that everyone else overlooked. (That can happen with
real paper reviews, too.)

Some questions may be controversial and so are good for
discussions. E.g., is the paper important? Are the contributions
significant? Does it have real applications?  Did people like it
(or hate it)? Opinions will often vary widely.