Joshua Steiner's Practicum Reflection Essay

For my practicum, I became an undergraduate research assistant for an electrical and computer engineering professor here at UMD. My work did not have a physical site, as it was a virtual position however, we are based at the University of Maryland. When first thinking of what I should do for my practicum I thought that trying to become a research assistant would be a good choice since I would definitely learn a lot , gain some connection with a professor which could benefit me come time to apply for grad school, and figure out if what I was doing seemed interesting enough for me to pursue a career in some related topic or field. Once I decided I wanted to do research for my practicum I started paying attention to the weekly email newsletter that the electrical and computer engineering department sends out and I would look for any advertisements for undergraduate research assistants. Not long after I found an advertisement, applied, and not long after completing an interview I was offered the position. If you are interested in becoming a research assistant my best advice is to find what kinds of topics you are interested in, and then research professors and email them about if they have any open research assistant spots and talk about how their work in 'blank' interests you and explain what your interests are. If you also know where their office is you could also stop by and ask them in person, as this would definitely make a good impression and I have a personal friend who did this and landed a research position that exact way. You can also do what I did which is looking at weekly emails from your department, but if you are interested in research work in a different department you will want to resort to reaching out to people. Ultimately you have to reach out to people and just express your interests and why you want to do the job. For my position, I performed assistant tasks with regard to working on an AI-machine learning model for human detection in conflict/disaster zones, using Unmanned Aerial Vehicles (UAVs/drones). I have been in charge of data collection of drone videos from the internet so that the model being built can perform successful human detection on real individuals in complex poses and chaotic scenarios (war zones, earthquake aftermath, etc…). I perform annotations of these images I find indicating where people are located in the image and marking what types of poses the individuals are making (standing, sitting, etc…). As I have worked on this, I have learned how AI-machine learning is only as good as the data provided, and for scenarios like the ones we are trying to deal with, there are not a lot of data sets out there. In order to start doing something like this we must build up a sizable data set of chaotic scenes with people in them. Also, I have learned how there is work being done with synthetic data to help build up detection models (synthetic being like little video game-looking characters in unreal engine). The field has a lot of innovation going on right now. In order for me to do this work I have had to do a lot of research on not just AI-machine learning topics, but also on the world. Specifically on what types of events are going on. Our data is dependent on conflicts that occur and natural disasters that happen. To find the best quality videos from drones it is best to use current events so paying attention to the news is something that I have started to do. And additionally, by looking into some topics that I have found drone footage from, I have been able to learn new information and events that lead me to more drone footage and therefore allow me to add more to our data set that we are building. I have grown an appreciation for the people that build up these object detection and machine learning datasets as their applications are incredibly useful, but it is a lot of work and tedious labor to get everything done. This experience has encouraged me to continue doing research with this professor and helped confirm with me that I want to specialize in Signal Processing and Communication Engineering, as this research had some related topics to the specialization and I found them very interesting. This will probably lead into my career path as well, since I am intending on pursuing a career in this field of electrial/computer engineering.Overall this has been a positive experience where I have learned a lot, met great people, and has allowed me to figure out (at least a little bit), what I want to do in the future.