In the last and final week, I worked on my final poster and paper. I used Microsoft PowerPoint to put together my poster and I just used Microsoft Word for my paper. It felt outstanding to see all of my work finally come together. Overall, I truly enjoyed my experience doing research in the Computer Science department at the University of Alabama. I learned so much and would love to do it again!
On Monday of this week, we got to meet up with the SITE (Student Introduction To Engineering) Program here and talk to high school students about computer science topics, courses, and salaries and also go the chance to let them learn about the research projects we were doing.
Afterwards, we got to visit the 3D Lab on campus! There, we got to see the many different types of 3D printers available for engineering students and about some of the past projects the students had made in the lab.
For the rest of the week, we finally tested all 50 of our voice samples. There were a different range of voice pitches and accents, so it was really interesting to see how accurate each of our respective programs were. In the end IBM BlueMix was the most accurate, Windows Cortana came in second, and CMU PocketSphinx came in third. We formed our results, and from them, we made a short abstract to submit to the Grace Hopper Conference.
This week, I was able to get my app connected to the avatar! Now when I say the phrases, "Turn on the light", "Turn off the light", "Turn on the dimmer", and "Turn off the dimmer", the avatar will respond with the proper request.
This week, I continued to work on my speech recognition app. I added the alias' "front", "back", "patio", "side", "bedroom", "bathroom", "kitchen", "upstairs", and "downstairs" to my code to start getting more specific with my functions. Dr. Anderson also suggested that the three of us record 50 different people (between all three of us) saying three different phrases for the smart home (such as "Turn on the light", "Turn on the front light", and "Unlock the door") in order to test out our apps with different voices, accents, and ranges in tones.
This week, I finally got voice recognition code into my application. I added in some code that would compare an array of appliances (light, fan, dimmer, etc.) to the sentence spoken into the microphone in order to find out what the user is trying to do.
This week, we learned how to use App Inventor to teach high school students how to code. Although we did not get to teach them, it was very interesting to see how the App Inventor program works. After that, I continued building my voice recognition app.
Coming to the University of Alabama as a student from Auburn University, you could already guess that I had some mixed feelings about coming here, due to our historic football rivalry. However, Dr. Anderson greeted me with a warm smile and I immediately felt welcomed. As I entered the lab for the first time, she showed us a few of the projects that her graduate students had been working on including complex programs, intricate robots, and also showed us the Avatar system that she was working on herself. Then, I learned that for our project, we would be working on an app that would be accessed and used with Cortana, Microsoft's voice-controlled virtual assistant. Throughout this week, we learned how to use GitHub, became more experienced with Windows development in Microsoft Visual Studio, watched videos from Microsoft Virtual Academy in order to become more efficient at using C#, and learned more about the pro's and con's of using Cortana.