cv

M.Sc

George Mason University, Fairfax, VA, USA

Computer Science

PhD

George Mason University, Fairfax, VA, USA

Computer Science

B.Sc.

Miami University, Oxford, OH, USA

Computer Science

Research Engineering Intern

Johns Hopkins University

Developed reinforcement learning infrastructure and robotic control systems at the Institute for Assured Autonomy.

• Led team adoption of NVIDIA Isaac Sim/Lab, building GPU-accelerated RL training infrastructure supporting 10,000+ parallel environments with vision sensors for Unitree Go2 quadruped navigation
• Designed bilevel hierarchical control architecture separating high-level navigation from low-level locomotion, trained using deep reinforcement learning (PPO) in PyTorch and deployed on physical robot via ROS2
• Enabled Matterport3D indoor environments in Isaac Sim by developing automated mesh processing and lighting injection pipeline to convert incompatible simulation platform assets
• Implemented multi-sensor suite (LiDAR, IMU, RGB camera) matching physical robot specifications and designed observation spaces and reward structures for neural network policy training
• Set up AprilTag localization system throughout lab to enable vision-based odometry correction for mobile robots

Graduate Research Assistant

RobotiXX Lab, George Mason University

Conducting research in robotics, motion planning, and machine learning for autonomous systems.

• Building neural network-based contact prediction system to enable failure-aware motion planning for home service robots in dynamic environments
• Implemented motion planning algorithms and collision checking interface in MuJoCo, improving experimental efficiency and enabling rapid prototyping for research team
• Developed and deployed ROS-based autonomous navigation and manipulation system for mobile manipulators (Fetch, Spot) supporting multi-room task execution
• Engineered open-source wearable egocentric sensor suite integrating multiple modalities (RGB, depth, IMU) for large-scale human motion data capture
• Led collection and curation of MuSoHu dataset covering 100 km, 20 hours, 300 trials, and 13 participants across diverse public spaces, resulting in IROS 2023 publication

Research Assistant

Miami University

Conducted research in deep reinforcement learning for UAV systems and machine learning for behavioral data analysis.

• Designed multi-agent deep reinforcement learning framework for cooperative UAV control under communication constraints, resulting in Springer book chapter publication
• Built machine learning models achieving 90% accuracy for behavioral data classification and developed interactive visualization tools using NetworkX and JavaScript

Technical Lead

Google Developer Group at GMU

Lead organizer and moderator for university seminars and annual Google Developer Group events at George Mason University, conducting live interviews and leading hands-on workshops on Google Cloud and applied machine learning tools.

Finalist for Best Paper Award

AAAI Fall Symposium on Artificial Intelligence for Human-Robot Interaction (AI-HRI)

Recognized as a finalist for best paper at the AAAI Fall Symposium on AI-HRI, highlighting exceptional research contributions in human-robot interaction.

VEF 2.0 Fellowship

VEF 2.0 Program

VEF 2.0 Program: Supports talented Vietnamese students in graduate studies at top U.S. universities and promotes U.S.–Vietnam educational exchange.

Human Uncertainty-Aware MPC for Enhanced Human-Robot Collaborative Manipulation

IEEE 7th International Conference on Industrial Cyber-Physical Systems (ICPS)

Developed model predictive control approach that accounts for human uncertainty in collaborative manipulation tasks, improving safety and efficiency in human-robot interaction. Co-authored with Al Jaber Mahmud, Filipe Veiga, Xuesu Xiao, and Xuan Wang.

Anticipatory Task and Motion Planning

Preprint

Research on task and motion planning systems that anticipate future states and requirements for autonomous robotic systems. Co-authored with Roshan Dhakal, Tom Silver, Xuesu Xiao, and Gregory J. Stein.

Toward Human-Like Social Robot Navigation: A Large-Scale, Multi-Modal, Social Human Navigation Dataset

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

First-author publication presenting the MuSoHu dataset, a large-scale multi-modal dataset covering 100 km of human navigation data across diverse public spaces to enable more natural social robot navigation behaviors. Co-authored with Mohammad Nazeri, Amirrera Payandeh, Aniket Datar, and Xuesu Xiao.

Responsive Regulation of Dynamic UAV Communication Networks Based on Deep Reinforcement Learning

Broadband Communications, Computing, and Control for Ubiquitous Intelligence. Wireless Networks. Springer

Applied deep reinforcement learning to optimize communication network regulation for UAV swarms under dynamic conditions and communication constraints. Co-authored with R. Zhang, M. Wang, L. X. Cai, and X. Shen, resulting in Springer book chapter.

Mapping the Complexity of Suicide by Combining Participatory Modeling and Network Science

IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining

Applied network science and machine learning to map relationships between suicide-related concepts, achieving 90% classification accuracy. Developed interactive visualization tools for exploring causal relationships in behavioral data. Co-authored with Phillipe J. Giabbanelli and collaborators.

Energy-Based Diffusion Language Models for BabyLM

Developing Energy-Based Diffusion Language Model (EDLM) that combines discrete diffusion with sequence-level energy functions to improve sentence well-formedness on the BabyLM benchmark dataset.

• Implementing compact diffusion model with contrastive learning objectives on 10M word tracks
• Comparing performance against baseline autoregressive and standard diffusion models
• Applying energy-based modeling to resolve failures of standard diffusion models in capturing global sentence coherence

LinkedIn Job Posts Tracking

Built an end-to-end ETL pipeline and recommendation system for tracking and analyzing LinkedIn job postings using cloud infrastructure and machine learning.

• Built ETL pipeline that scrapes LinkedIn job posts into MySQL database on AWS EC2 instance
• Designed S.T.A.R schema database to store over 1000 job postings using MySQL and Python
• Extracted job details and requirements using regular expressions and Named Entity Recognition (NER)
• Delivered personalized job recommendations by aligning users' resumes with job postings