1a) As mentioned in my last progress report, I've set up some branding elements (Twitter account, Website), and set up version control. I've also refined my hand-tracking system to have smoother animation and a truly 'still' cursor when your hand is reasonably still. I'm currently in the process of implementing relational tracking for the hands. Immediately following that will be implementing tracking for both the active hand and the inactive one to allow for a smooth cursor handoff, for lack of a better term.
2) How many times have you been in contact with your mentor? List what you've discussed.
2a) I've met with my mentor twice, briefly since our first formal meeting where she approved my timeline and proposal. We discussed version control solutions and setup, and she also helped me write the simple algorithm I'm using for animation smoothing. We plan on meeting on Thursday, where I hope to discuss the relational tracking and help build my immediate timeline.
3) How has your timeline changed thus far? Discussed what changed and why you changed it.
3a) Primarily, the biggest change has been pushing back my goal-lines for this semester. I had hoped to have a very basic prototype done, and I'm still not there yet. After the goals mentioned in the answer to question 1, the last step before I feel ready to start writing the actual application is going to be gesture tracking, but that's pending my discussion with Dr. Baker. We may decide that it's worth putting on the back-burner between getting a rough interface written and the beginning of next semester, assuming that getting the interface isn't very fast.
4) What do you still need to complete in preparation for 499? Identify key aspects that need to be taken care of before your capstone starts.
4a) To itemize and elaborate on 3a a bit more, here's what HAS to be done:
- Hand tracking (both hands, with smoothing): tracking both hands is necessary for both the smooth transition for the cursor as well as leaving me a hook to assist in gesture tracking.
- Interface building: Right now, the 'application' is an empty canvas to move a cursor in. At the very least, I need to build a testing ground for different bounding areas and time-to-select so I can test those.
- ** Gesture tracking: I'm envisioning this being complex, assuming there aren't extensive libraries already available for use, so this is pending discussion. When hand tracking is complete, I will have the skeleton on which to add the meat of gesture-tracking, but having gestures will also require having an interface to gesture through, and writing my own gesture solution may require more calculations which may negatively impact performance.
- Coming up with a good solution to bring the Kinect with me: It's about as simple as it sounds. The Kinect is relatively fragile, and I need a good way to be able to bring it to campus with me. It may just be as simple as wrapping it in a towel or something and putting it in my bag, but I'll just need to find a way to prevent any pressure that may strain the motors.
5) If there have been significant changes to your project since your proposal, please discuss them.
5a) Nothing in particular. The biggest thing is I'm only planning on using 1 Kinect as opposed to the potential of 3.
No comments:
Post a Comment