Yesterday, my senior design team and I had demonstrations for our project. For those of you who don't know, our project was to develop a personal "smart mirror". We basically created a proof of concept for a mirror that you would display in your bathroom (or where ever, really), and it would greet you, and give you information about things that might be relevant to someone as they're getting ready for their day, such as weather forecast, upcoming calendar events, news headlines, and traffic updates for getting to work.
Although we didn't accomplish every single thing we set out to do initially, we were satisfied with what we accomplished and pleasantly surprised at the positive response we got from everyone that came up and talked to us about our project.
The application itself was just a responsive webpage hosted on an Apache server. All the information we wanted to display was relatively easy to do with various APIs and feeds, as well as HTML5 geolocation. There's also a configuration page you can surf to to set up your account and configure your settings for you calendar, work address, etc. As for the mirror itself, it's simply a LED TV with the frame disassembled, and a piece of glass with one-way mirror film attached with aluminum framing. I'm not going to go into too much detail, but you can read the project overview in the poster I made (which I'm quite proud of. It was unique and we received a lot of compliments for it :-).
As for the things we weren't able to accomplish...
Our initial idea was to set up facial recognition so that when a person approaches the mirror, the person would be identified so that their personal info could be loaded. We ran into memory issues on the Odroid while trying to run the facial recognition algorithm. By the time we figured out the issues, it was too late to switch out the board. We also had an issues with there being a quota with the speech recognition API if we ran our app in Chromium (which is why we ran it from my laptop during the demonstration), but Chrome isn't available for ARM processors. As they say, hindsight is 20/20, and we should've used an x86 board from the beginning. Lesson learned.
Now that our demos are over, I can finally relax. With no finals standing in my way, I eagerly wait for graduation next weekend! Next month, Stephen and I move into our new apartment in Plano, followed by a 3 week trip to Seoul and Tokyo before we start our new jobs in August. I'm so excited for what the future holds and I'm glad I have someone I can share all of it with. :)
Thanks for reading,
Christina