The Final Project is an opportunity for you to apply what you have learned in class to a problem of your interest! Here’s some logistics:

  • Theme: The project can be on anything that can use deep learning as a helper/solver
    • Project Type 1: If you’re coming to the class with a specific background and interests (e.g. biology, engineering, physics), we’d love to see you apply deep learning techniques to problems related to them! Note that this does not need to directly involve computer vision in any way. Additionally, the students can also choose to do a mini-literature review of a computer vision task that was or wasn’t covered in class.
    • Project Type 2: In this case, the students will choose at least two papers on that task, run their available codes and compare their performance. They will also a mini-introduction what that task is and how people have approached it besides deep learning, if ever. In the case that you don’t have any project idea for either type, the professor can also suggest project ideas and will be available to quick routine meetings if with team finds it appropriate.
    • Project Type 3: Here, the students will develop a software (a web app, an executable, etc.) that applies some of the deep learning and/or computer vision contents covered in class (the students are also free to explore other AI concepts). The software should be developed using concepts of MLOps and should be aimed to be used by a non-expert in the field (like a standard human user). This type of project will be more heavily graded on its application of software engineering protocols, its user friendliness and it’s code organization/documentation.
  • Team Formation: The teams should be of 2-4 people and they are formed at the discretion of each student. In case you are having trouble finding project partners, consult the professor.
  • Project Proposal: The teams should be formed by Dec 4th, when they should submit a 1-2 page project proposal. It should contain the problem statement, motivation, the main tasks and how each student will contribute to it.
  • Project Presentation: The final presentations will take place on the week before the finals period and each presentation should last for at least 10 min, such that each student member presents for at least 4 min. In the presentation, the team should introduce and motivate the problem, describe your solutions and the difficulties found and present some results along with the summary.
  • Deliverables: Each team should send two main deliverables: (1) your project code in a .zip file and (2) your slides. Your code will be mostly graded based on its organization, where documentation will play a big role, and on correctness (i.e., you code runs well and produces the expected results). Your presentation will also be graded on its organization (from both the slides and speech perspective) and whether it covered the topics described above.
  • Grade Breakdown: Proposal – 10%, Code – 40%, Slides – 10%, Presentation – 40%.