CSE 455: Final Project

Key Dates

Assigned	Thursday, Oct 16
Proposal due	Thursday, Oct 23 (11:59 P.M.)
Final reports due	Monday, Dec 8 (11:59 P.M.)
Final presentations	Wednesday, Dec 10 (4:30 P.M. – 6:20 P.M.)
Late due date	Up to 2 days after the original due date

Synopsis

For the final project, we anticipate that people will work in teams of three or four. There are two options for this project, which will take roughly four weeks:

Implement a state-of-the-art research paper from a recent computer vision conference or journal (CVPR, ICCV, ECCV or vision parts of NIPS, ICML, SIGGRAPH, SIGGRAPH Asia, etc.).
Complete a short research project (more fun!).

You can devise your own project from scratch, or use one of the ideas suggested below. In either case, the purpose is to learn more about and get a feel for doing research in computer vision.

All projects must have a machine learning component.

Guidelines

Option 1

Start by searching through recent computer vision conference proceedings or journal articles, and choosing a paper that interests you. The premiere vision conferences are ICCV, CVPR, ECCV. The premiere journals are TPAMI and IJCV. Sometimes computer vision papers also appear in graphics/machine learning conferences such as NIPS, ICML, SIGGRAPH, SIGGRAPH Asia. We recommend starting with the most recent years, i.e., CVPR 2023, ICCV 2023, ECCV 2022, etc. Most of the papers and project web sites are linked online from here. You should select a paper that is appropriate for a four-week project. Our expectation is that you will implement the method yourself rather than using any code that is available online.

Option 2

Do a research project with some novelty—something unpublished. While we’re not expecting PhD-level work in four weeks, a team of three or four can get exciting results :)

Examples:

Extension of prior work (ideally implement the base method yourself).
New application of a known technique; evaluate performance.
New solution to an existing problem (have the solution by proposal time).
Pose a new technical problem and solve it (have both by proposal time).

Ideas suitable for final projects (choose variations or your own): face/pedestrian/vehicle recognition; tracking by detection; re-identification; product description (CV+NLP); license plate recognition; human pose/hand gesture recognition; instance recognition; cancer biopsy diagnosis. More ideas on Kaggle Competitions.

Requirements

Proposal

Each team submits a one-page PDF on Gradescope including:

Team members
Project goals (input/output)
Milestones

Add your project title and name of the team members to this spread sheet.

Final Presentation

End-of-quarter session: each group presents (PPT over Zoom).

Final Report

~10-page PDF on Gradescope. Include each member’s contributions (end of the introduction is fine).

Use the NIPS format. You can use the .docx or LaTeX (.tex + .sty). Overleaf is a great collaborative LaTeX option.

Your report should include: title & team, short intro, related work, technical description/algorithm, experiments, discussion (strengths/weaknesses), and future work.

Grading Rubric

Oral Presentation: 10 pts
Quality of Report: 10 pts
Experiments and Results: 15 pts
Creativity and Novelty: 10 pts
Total: 45 pts