The most significant risks that could jeopardize the success of the project is that, based on professor feedback from our proposal presentation, we have decided to use the NVIDIA Jetson Xavier NX Developer Kit instead of the Nano as previously decided. The biggest risk associated with this is the cost, namely that it costs $399, which is significantly higher than the Jetson Nano which was $99. This eats into our overall budget pretty heavily and does not allow room for any mistakes to occur. If, for some reason, the Xavier were to somehow become usable, we would have to re-evaluate how to proceed with our contingency plans. To lower the risk of damaging the Jetson Xavier in some way, all of us will educate ourselves on how to properly handle it. We will also be storing it at Danielle’s apartment since the carpet in Heidi and Vaheeshta’s apartments pose a significant ESD risk, as well. Our current contingency plan is to try to find a cheaper, used Jetson Xavier option on a resale website such as eBay. We will discuss with our TA, Abha, about any other suggestions that she may have for contingency plans.
We have made several changes to the existing design of the system, following our Project Proposal presentation from Wednesday. We met to take into account the feedback that we received and think about how it would affect our project. As stated above, we have decided to use the NVIDIA Jetson Xavier NX Developer Kit instead of the Nano. We have also decided to adjust the scope by removing eye detection as one of our requirements, since it combines with facial detection, and instead using DLib 68 points landmarking. We also intend to look into adjusting our landmarking accuracy by researching more based on Professor Savvides’ comments following our presentation.
None of our changes have affected the schedule of the project, so there are no updates.