Week 10: Pixels and Practice

Hello, and welcome back to Week 10 of my senior project blog! With Commit Day finally behind me, a huge weight has been lifted off my shoulders, and I now have the freedom to focus fully on preparing for my capstone exhibition. This week’s major victory has definitely been completing my presentation slides, and the primary hurdle I had to overcome has been ensuring that I don’t bore my audience to death with dry TensorFlow metrics and technicalities. Instead, I decided to craft a compelling narrative: a story of classical AI’s inability to cope with the chaos inherent to real-life data, such as blurry dashcam footage, and how QNN entanglement acts as a natural buffer against errors.

While crafting my presentation, I have also been busy working on my exhibition poster. I am almost finished with it, but one final element still needs some work. At the moment, I find myself struggling immensely with the visual hierarchy, specifically placing two contrasting sets of clear, static-filled traffic light images precisely next to the degradation curves to highlight the difference in performance between my QNN and its classical counterpart. With only a couple more tweaks to the poster design, it’ll be ready for printing.

I worked hard to make sure my visual deliverables look great. However, the true horror of capstone presentations is finally dawning on me. I actually have to verbalize and explain all of this content within a limited time to a live audience. While playing with circuit designs in Google Colab is certainly much more relaxing than presenting them aloud, I will be doing the latter next week.

Patrick Zhou, signing off.