Throughout the course of the semester, we will read a number of papers on several different topics. Each of you will be designated to prepare written notes on one of these topics. Nominally, your in-class presentation will be on one of the papers in the topic you have been assigned.
At the beginning of the semester, each of you will be assigned a particular topic (e.g., mesh compression). Most topics will generally cover two lectures.
Prior to the relevant lectures, I will post the topic outline on the class schedule page. These outlines will include information on required papers, recommended papers, useful links, etc.
At the beginning of the class discussion of your topic, you should turn in to me an outline of the survey you plan to write. This outline should indicate the overall structure of the paper, key ideas you plan to highlight, and a list of papers that you will discuss. Make sure that I approve this outline.
During the week, you will presumably give a talk on one of the required papers.
The following week, you should turn in to me the draft of your survey. This is what your grade will be based on. I will look over it, and may suggest changes.
The final draft of your article, including the revisions I request, will be posted online.
To hand in your paper, you should e-mail to me both the source files for your article and a PDF file. The “source” files are whatever is necessary to make the PDF file. If these files exceed 2 MB, please put them on the Web and send me a URL instead of e-mailing me the files themselves.
Here are some general guidelines on what I expect to see in the notes you’ll be writing.
You should begin by defining the problem setting. What is the problem being addressed? In what applications does it arise? What is its cause? What kind of solutions are we seeking? What are we trying to accomplish?
Having set the stage, you should provide a survey of the most relevant research papers. This must include the papers assigned as “required reading”. It should also papers listed as “recommended reading”. Any other particularly salient/interesting papers should also be included. Provide a high level summary of each paper. Mention its key ideas. Give some indication how it relates to the other papers you are surveying.
Include any material from lecture that you consider important. Standard mathematical formulations, proofs, algorithms, diagrams … anything that would help a reader more fully understand the literature you’re surveying.
Almost every paper we will read — indeed almost every research paper published — contains a survey section. It is typically entitled Related Work, Prior Work, Background, or some such similar thing. Read these and study them. Pay attention to how they’re written. There are many articles written as surveys of a particular field. Here are a few examples of such survey articles that are particularly relevant to the material in this course:
P. Heckbert and M. Garland. Survey of Polygonal Surface Simplification Algorithms. In SIGGRAPH 97 Course Notes: Multiresolution Surface Modeling.
M. Garland. Multiresolution Modeling: Survey & future opportunities. Eurographics ‘99, State of the Art Report, September 1999.
This is exactly the kind of writing that you should strive to produce. Notice that they all provide a well-thought out description of the problem, and discuss many relevant papers. The surveys you’ll write will probably be significantly shorter, but their goals will be the same.
Producing an aesthetically pleasing document requires the right tools.
For the purposes of this course, there are two acceptable systems for typesetting your documents:
Microsoft Word. This is barely acceptable, but I’ll tolerate it. The mathematics produced by Word and its hence-program, Equation Editor, are simply ugly. I don’t want to be forced to look at this any more than necessary.
LaTeX. This is the tool that I strongly recommend. Yes, I know that TeX is a HaX. I could easily write a fairly long screed on everything wrong with (La)TeX. But to paraphrase Winston Churchill: LaTeX is the worst typesetting tool we have, except for all the others. Anyone seriously considering Computer Science research must learn LaTeX. So if you don’t already know how to use it, do yourself a favor and learn.
If you need to install LaTeX on your machine, here are some quick hints on how to get it: