This week was crazy. During our meeting, Karsten pointed out to me that I should have a backup plan in case my classifier is not ready by the time I do the presentation. He advised me to make a small visualization of the recorded signals and corresponding features (i.e. statistics), after which I could still try to build the classifier.
Since I was able to smooth the signals now, it was fairly obvious how to approach this visualization challenge. I started looking for some charting libraries and quickly found that MP Android Chart is currently the best one out there. To calculate the corresponding statistics, I decided to make use of an Apache library called Commons Math.
So yeah, that’s basically what I did for an entire week: making a pretty interface to present those signals and stats. Instead of overwhelming you with screenshots of every single step of the execution flow, I will attempt to make things clear in this demo video of my Android prototype application.
After having implemented this basic visualization, I started on my presentation that was due to Tuesday morning (and not Monday, as I wrote in my previous blog post). For those who are interested: the final Powerpoint presentation, in which I basically explain my entire thesis, is available here.
Today, one day after the intermediary thesis presentation for my colleagues and a large part of the HCI department staff, I am happy to inform you that it went very well! Even my live demo (the video above was a backup plan) proved to be a great success.
So why was this week crazy? I worked for about 60 hours on my thesis alone in order to make the deadline (which I eventually did). Having said this, I think I’ll allow myself to take things a bit slower this week: I really need the sleep and, more importantly, the time to work on my other projects…