Ill Tidings
It’s been such a strange time. It feels as though 2020 has taken another drastic turn and this time eyes turn to Texas in the tale of the COVID-19 infections. I cannot say I am not scared, even if not lethal, the damage is very real for those who exhibit symptoms. I have already had two scares now in the span of less than 3 weeks. It’s hard to focus on things and hunker down when now my own mother has been exposed because she works in the Hospital system here in Houston. By Tuesday, I will be back in my new apartment in Austin, away from family, away from this. It feels odd but, that’s not supposed to be the focus of this. I did learn and do some good things, despite how this opener might sound. There continues to be hope and small reprieves.
Gaze Detection & IDE Problems
As we dove deeper into the world of machine learning, we began to discuss gaze frequently as it pertains to being able to track movement using a camera. One of our guest speakers this week, Anil, presented upon how his research was trying to more accurately verify whether or not people were actually looking at a screen, tracking screentime. There were different classifications of “accuracy”, however, according to Anil and his research the majority of the time, the machine was able to track whether or not there was someone looking at the screen. He had mentioned that one of the issues they had run into was when there are multiple people in front of one screen, pending position. This addressed a question that I had from one of the readings we had done, “Appearance-based Gaze Estimation in the Wild (MPIIGaze)”: is one of the difficulties that presents itself largely to do with the fact that the machine struggles with depth? I couldn’t make much sense of it, but looking at how it converted 3D images into something that was 2D, I could see how data may or may not be “lost” when doing the 2D rendering. And, greyscale is used versus the RGB colors in order to make for a more efficient rendering it seemed like. The dimensions for converting from 3D to 2D were still odd to me and when I asked Anil if that could’ve been a source of error as well, he did give some credence to the thought.
One of the tasks that we had to perform this week had to do with installing Open CV and using an IDE in order to work with that. This became the thorn of my existence and when I tell you, at one point, I was about 3 seconds away from becoming familiar with a distant distilled Russian friend of mine, I do mean it. We had settled on PyCharm and for whatever reason it was not working for me: maybe I didn’t download the packets in the right form, maybe I had the wrong version of Python, maybe I was using the wrong bit-size version of it. Mary Jin gave some great advice on how to proceed but, unfortunately, that’s not before she got to see the monster that I had been letting fester.
Everything went wrong
Needless to say we have switched over to Anaconda and things are much nicer and smoother now. Everything is finally downloaded so I’ll be playing catch up with making sure I can create the boxes around the eyes for gaze tracking. But, I feel more confident now.
Things to Come
The things that I have been thinking about so far that will require the most brainpower from me is thinking of how to create the lesson for my students from this material that we have been tearing through. Jimmy, one of our other distinguished presenters, was extremely insightful as to what should be expected of us when we submit our lessons to TeachEngineering. He was also very clear that some of the things I was experiencing anxiety about were not to be that big of a deal. Now, the scripting and pacing & the student work will be the thing I need help with once we establish teams and work on generating our lesson plan. I truly want to start working on something social media related, but understand if that’s something that would miss the mark or has already been done.
I am having some reasonable self-doubt about my ability to pour my all into this and fulfill my other roles right now. That’s not to say that I don’t want to participate in the program, rather, I just want to make sure I’m doing right by the university and doing right by me. I want to learn, I want to do better. This Machine Learning course by Andrew Ng and the other additional materials need to get examined and reviewed not because I need to have them done, but because I want to understand it all. These past two weeks have made me re-examine my current trajectory within education and have me thinking maybe this could be something I explore further as a professional career down the line. And, continued shout-out to my cohort, they have been a great sounding board and I am glad that they sometimes give me reason to pause and say, “You know what, they’re right.”
Research Buddies
