Defending my research of 1 1/2 years in video analytics on the MLK Smart Corridor Testbed.
Accelerating H.264 decoding overhead using the Pipes and CUDA-enabled FFMPEG to bring a smooth streaming experience to OpenCV.
Using NOAA, OpenFEMA and hospitalization data to predict hospital surges based on natural disasters.
Over time, my mods have gotten better and have received more and more downloads, to the point that just a few months ago I officially hit 10 million mod downloads!
A 1-mile stretch of Chattanooga's MLK Blvd. dedicated as an infrastructure.
All-in-One Mapping Program to enhance driver awareness
See-through technology for vehicles using infrastructure.
Winners of the first-annual programming competition!
A Stock-Image Consolidation Program
On March 4th 2020, I defended my thesis called "Video Analytics on the MLK Smart Corridor Testbed" and passed! This thesis discussed many forms of real-time scalable video analytics that were done under CUIP. You can see a dashboard of this work below!
In 2018, Dr. Mina Sartipi at The University of Tennessee at Chattanooga started a research group in the SimCenter, the Center for Urban Informatics and Progress (CUIP). CUIP has aimed to be a nexus for collaboration with members of the city, other universities, and more. With our partnerships, we have been granted 16 poles on a mile-long stretch of Martin Luther King Blvd.. With this, we have launched a scalable infrastructure capable of expanding both hardware and software. As part of an ever-expanding project, I have created an object tracker (seen here) that is parallelized, resilient and produces nearly 500,000 data entries a week per camera. This algorithm allows us to store anonymous data on Kafka, and functions in real time. This algorithm allows for many data analytics in the future, and has already revealed many trends that can be addressed by the city or its members.
To follow up on our CAV (Connected Autonomous Vehicles) initiative, for the 2018 conference the team decided to work on a unified map for anonymously mapping many subjects of interest. For the most part, the above video I made should suffice as explanation, but it was fun being the only one working on the infrastructure end.
Without a doubt, we couldn't have done this without Andrew Rodgers, so a million thanks to him for helping us with our testbed.
This programming project was quite a hurdle for me. The 2017 SCCC was in late June, and I joined the research group in early May. I had never done networking, computer vision or machine learning in my life, nor had I ever written in Python. Yet our team managed to get this together, with myself handling at least 70% of the workload. This project had a live-demo that took everyone by surprise, but was incredibly stressful on my end. Nonetheless, this was the first time that I felt confident that I could accomplish anything I ever wanted.
"CECS Students also took top honors during the app development component of Startup week. InfoSystems and Skuid presented a $10K scholarship to grand prize winners Alay Patel, Jose Stovall, Steven Hullander, and Evan Grayson, who are all juniors in the Computer Science and Engineering Department. The fifteen teams in the competition were tasked with creating a mobile app that allows consumers to digitally keep track of their physical conditions. The first and second place teams excelled in creating an inviting and intuitive UX/UI design, but the UTC “Brogrammers” took it a step further by adding a mini-game to their app." ~Holley Beeland
I will not be revealing too much information for the sake of respect to the Nissan dealership which offered me the program contract, but this marks the first time I was ever paid to program! The program was a simple, UI-based string comparator which allowed a user to select a file, paste a list, and then compare both string collections and output which did not match. While this sounds vague, consider it in the context of comparing incoming vehicles with current stock.