Prithvi Raj Singh

About

I am currently a Visiting Lecturer in the Department of Engineering & Computer Science at McNeese State University, where I teach programming fundamentals and machine learning. My work bridges theoretical computer vision research with practical applications in robotics and edge computing.

I completed my MS in Computer Science at the University of Louisiana at Lafayette under the co-advisement of Dr. Raju Gottumukkala and Dr. Anthony S. Maida., focusing on real-time detection and tracking of fast-moving small objects tailored towards Racquetball playing robot. I completed all my research work in the Department of Mechanical Enginnering. Prior to that, I earned my BS in Computer Science with a minor in Mathematics from McNeese State University, where I served as President of the ACM chapter and founding member of the robotics club.

Originally from Birgunj, Nepal, I grew up in a working-class family and spent much of my childhood in my ancestral village. After completing my high school, I came to the United States to pursue undergraduate degree with plans to get PhD. I am proud of how far I have come.

Research Interests

My research focuses on advancing 3D computer vision algorithms for robotics applications, particularly in challenging scenarios involving fast-moving small objects. I work at the intersection of classical computer vision, deep learning, and physics-informed models to create robust perception systems that operate efficiently on edge devices.

Primary Areas

• 3D Computer Vision
• Robotics Systems
• Scene Understanding and Reconstruction
• Object Tracking & Detection
• Robotics Perception
• Edge Computing & TinyML
• Reinforcement Learning

During my graduate work, I developed a hybrid system that use kinematics motion model along with YOLO detector to detect and track tiny racquetball. This work involved extensive experimentation with various detection architectures (YOLOv5-8, SSD, Faster R-CNN) and tracking algorithms (Kalman Filter, SORT, Deep SORT) to handle motion blur and occlusions. The results demonstrated significant improvements in tracking accuracy for fast-moving small objects in 3D space.

Teaching

At McNeese State University, I teach courses spanning from introductory programming to advanced machine learning:

Professional Experience

Software Developed

GeoTechPileCPT — Designed for the Louisiana Department of Transportation under Federal Highway Commission funding. Used for geotechnical analysis and pile foundation design.

Contact

I'm always open to discussing research opportunities, collaborations, or PhD positions.