Harvey Mudd College
Computer Science 153
Final Project

Important Deadlines

Final Project Overview

The final project is a project of your choosing on any topic in computer vision. You will work on this project in teams of 2, 1 or 3 (with 2 being strongly recommended). This project is expected to be an in-depth look into a particular aspect or field of computer vision. It should go into substatially more depth than any of the problems sets have.

Your project can address any field of computer vision and can take a number of differernt formats. The following are possible (but not the only) formats for acceptable final projects:

• A solution to a specific problem of your choosing (for example, recognizing sign language, visually keeping track of your textbooks in your dormroom, recognizing poker hands, etc), using an existing algorithm or a new algorithm of your own design.
• An implementation of a previously published algorithm.
• A detailed survey paper of a particular area of computer vision (note: if you choose a survey paper, expect to do a LOT of reading and research).
• A new algorithm that is designed to improve on a flaw of an existing algorithm for some problem in computer vision.

An open-ended project is likely to be too difficult for the time you have, so I would suggest narrowing the scope of your project, for example by limiting the cases that you would like your solution to work for, or by making other simplifying assumptions about the domain (e.g. lighting is constant, etc).

If you are having trouble choosing a topic or topic area, look back over the possible suggestsions listed on the problem sets for inspiration, or come talk to me. There are so many great problems in computer vision!

Final project requirements

• Project Proposal
  

  For details, see below. Project proposals are due Friday, Oct 28.

• Meeting
  

  At some point in the two weeks before the end of classes, each team should schedule a meeting with me to evaluate progress on the final project. To help with this scheduling, the class time on November 23 and 28 will not be used as lecture or lab time, but reserved for meetings or work on the project.

  At the meeting, plan on demonstrating some piece of your system (unless you are writing a survey paper, in which case you should know a significant amount about the field.) Also, it will give us a chance to evaluate the overall scope of the project and discuss ways of progressing from its current state.

• Write-up
  

  The write-up is the formal documentation for the project. I would like you to think of this as a real (potential) submission to a conference or workshop. CVPR (Computer Vision and Pattern Recognition) is one of the main computer vision converences. Although their deadline will pass before you have your paper done (Nov 7), we will be using their template for formatting our papers. It is unlikely that you can complete enough work to have a submittable paper in this short time (most people spend at least a year or two before they have enough to submit to CVPR!). But if you get excited about your project, I would encourage you to continue working on it after the end of the semester and plan to submit it to CVPR next year. Your paper must adhere to the CVPR template, but I am imposing a 5 page limit, instead of the 8 imposed by the conference. You can download the necessary tex and style files from the CVPR submission site. Please see me if you have any questions about this.

  As far as the sections you include in the paper, there's no standard format, but the following sections outline a common approach to covering the important points:

  • An Introduction that motivates and describes the problem and the results at a high level.
  • A Background section that explains the basics of the algorithms you applied to the problem.
  • A System Description section that provides the details of how you constructed your system, how it works, and how you tailored the algorithms described in the previous section to the problem at hand.
  • A Results section that describes how well the system performs. A format that often works well here is to first explain your evaluation techniques, provide their results, and then explain those results and what they say about the problem and about your approach(es) to it.
  • A Conclusion, usually very brief, in which you can summarize your system and the results. In addition, this is a chance to be less scientific in your opinions about the project and a chance to put it in the larger context of larger, more general problems (such as the general vision problem or a broad subfield).
  The write-up should not include a print out of code, but should indicate where that code (and, if applicable, test data) is available. The write-up is due by 5pm on Sunday, Dec 4.

  Following the paper submission, there will be a 2-day review period, during which time you will be required to review 2-3 of your classmates' papers and give them feedback. These will be similar to the reviews we have been doing all semester. Reviews will be due on Wed, Dec 7 by class time.

  At the end of this review period, you will make revisions and submit your final paper by 2pm on Monday, Dec 12, at which time you will also give your final presentation (see below).




• Presentation
  The presentation is a chance to summarize your project and results in a compact, semi-public forum. It need not be long (10-15 min.) and should include roughly the same components as the write-up, along with a demonstration of your system. The presentations for all of the projects will be schedules for the final exam time: Monday, Dec 12 at 2pm. However, you may schedule it earlier if you like.

Final project proposal

• Overall project description
  What you'd like to do and how you can evaluate it. You should consider the scope of your project here -- in particular, suggest ways that the project could be limited or simplified, if need be, as well as some ways in which it could naturally expand.



• References
  Include at least one, but preferably three or more, references to work you've found that is related to your project. (If you choose to do a survey, you MUST include at least 3 references in your proposal). Consider the similarities and differences between that work and what you'd like to do (even if just based on initial impressions). How helpful do you feel each reference will be? Include the location of the resources (web site, book title, paper reference) (that way I can get more familiar with them...).


• System and Starting point
  Explain how you intend to start the project. Particulars of the start-up include the language (and perhaps platform) under which you anticipate working and the source for input to your system (still images and where they'll come from, or, perhaps, video from somewhere). Are there any hardware requirements that I need to be aware of (so we can start setting things up)? Describe in as much detail as you can a small initial piece of the project that you intend to start with. Also suggest a number of separate pieces that will comprise the whole system: what are your initial thoughts on the order in which to attack them? Are there any that will be more difficult than others?