Undergraduate Research Assistant

I am currently involved in Prof. Mohammad Al Faruque's Research Lab.

During my first quarter, I was tasked with exploring the capabilities of the OpenUAV VLA model and conducting training and performance evaluations. I built several Python scripts on top of this platform to analyze datasets, format scene-level results, generate multi-angle first-person drone videos from sequential frames, and visualize 3D trajectories to compare model-predicted paths against ground truth trajectories.

During my second quarter, I contributed to the high-level implementation of our proposed design "AeroLANCE" on a physical UAV. I supported system integration, operationalized the pipeline on the drone, and conducted structured task-based evaluations across multiple navigation scenarios to validate its performance. I operationalized and validated our proposed AeroLANCE navigation architecture on a physical UAV through structured task-based evaluations across multiple navigation scenarios. I also reimplemented the TypeFly baseline within the same codebase to support controlled comparative experiments. In addition, I performed monocular 3D reconstruction from physical UAV image sequences using MonST3R and generated visualizations in Blender.