Presenting my Research at the Undergraduate Research Symposium

Author: Matt Couts | Major: Computer Science | Semester: Fall 2024

Hello, my name is Matt Couts and I am a recently graduated honors student from the College of Engineering, where I studied Computer Science. In the Fall 2024 semester, I conducted research in artificial intelligence and computer vision under the mentorship of Dr. Khoa Luu, a faculty member of the Electrical Engineering and Computer Science department. Since graduation, I have joined Walmart as a software engineer on a team that utilizes computer vision and artificial intelligence to enhance the Walmart shopping experience.

When I first looked for undergraduate research opportunities, I was drawn to the field of artificial intelligence because of my fascination with its explosive growth. After reading about Dr. Khoa Luu, his research, and his accomplishments on the faculty website, I reached out to him to express my interest in researching under him. Following a discussion about my interests and motivations, he agreed to mentor me and introduced me to the Computer Vision and Image Understanding lab he oversees.

In the lab, I met several incredible students, including Ph. D. students, who already had established research rhythms and routines I could mimic. One of my favorite parts of this research experience was hearing about the projects the rest of the lab was working on. Their passion for their research was both inspiring and encouraging to me.

Among the students in the lab, Pha Nguyen stood out and became a peer mentor of sorts throughout my research journey. He helped me select my research topic and provided me with relevant literature and technical guidance. I learned a lot from Pha, especially during the brainstorming sessions he and I had throughout my research.

Video language models are machine learning models that learn from both video and text data. These models have the potential to impact the real world in significant ways, making this topic particularly compelling. For instance, video generation models, which generate videos from text prompts, are already revolutionizing the world of content creation.

My research focused on video understanding, a branch of study that examines how well video language models can converse in natural language about the contents of a video. Specifically, I investigated how modifications to a video language model’s memory mechanism affected its ability to comprehend and respond to questions about a video. I measured the model’s performance using a video question-answering dataset, which contained a collection of videos paired with questions and their corresponding answerings.

After graduating this semester, I joined Walmart as a software engineer on the Store Vision Services team, a team working to improve the Walmart shopping experience using computer vision and artificial intelligence. The experiences I had researching under Dr. Luu in the field of computer vision and artificial intelligence was pivotal in securing this position and has completely shifted my career trajectory.

Looking back, I can confidently say that undergraduate research was the most transformative experience of my time at the University of Arkansas. It not only helped me grow as a student and researcher but also sparked my passion for innovation. I am extremely grateful to the Honors College Research Grant for enabling me to pursue my research interests.