Students will scope a potential application of computer vision of interest to them (ideally, applied to their research!). Projects can be done solo or with one partner without approval. If teams larger than two would like to form, you will need to obtain permission from us by submitting a written outline of what each group member's distinct role will be as part of the project scoping follow-up. In addition to the milestones described below, we will devote class time to workshopping project ideas, issues, and progress. Please come prepared to discuss!
The purpose of this first project milestone is to get you all thinking about these early, to spread and workshop ideas in class, and to help with matchmaking for group work. If you have already found a partner/group, you may pitch with that partner.
On Jan 31, each person or group will give short (2 min) pitches on one of these ideas. Please come with 1-2 slides! Ideas and groups need not be at all finalized. Feel free to post in Discussion on Canvas if you're looking to form a team. Please prepare slides according to the format guide and upload to Canvas.
By now, you should have narrowed down a project idea, and if working with a group, who is in that group (and what each person's role is, if 3 or more people). You will submit a brief (half to full page) writeup of your plans. This should give us a sense of background for the application, the data plan -- what existing data to use (e.g. "I have this dataset/I'll use this publicly-available dataset") or a plan to capture it (should be easy, e.g. "Over the next two weeks, I'll record X on my phone/webcam..."), and a computer vision task you want to try. If you are working with sensitive data, you should identify publicly shareable proxy data that will allow us to get a sense of what you are working on.
This will be submitted in Canvas, and we'll follow up with comments.
For each class, come prepared with a 5 min (Please time this ahead of time! It's not a long time.) presentation explaining the scope, progress, and current challenges. We will use these to workshop your projects in class. Please prepare slides according to the format guide and upload to canvas. This presentation should include (1) identification of scoping refinements and shifts, (2) a first demo in which you apply some computer vision technique (need not be your end application -- this can be super lightweight, e.g. using an out-of-the-box tool and display a possibly-not-working result), and (3) thoughts on where you would like to take the project for the rest of the quarter. Include figures!
Final presentations! Slide presentations --- introduce us to your project, describe what you did, and describe where you might like to go from here. Please upload slides to Canvas.
Please provide a final report, 4 pages. Please use our rubric as a guide.