cv

Research Intern

HKU AIoT Lab, The University of Hong Kong

Selected as 1 of 40 students for a fully funded summer research internship; supervised by Prof. Chenshu Wu on privacy-preserving fall detection using low-resolution thermal (62×80 IRA) sensing.

• Re-architected the real-time inference pipeline for NVIDIA Jetson Orin, raising throughput from 6.7 to 20.8 FPS and cutting display latency from 3 s to under 0.3 s, via a CNN feature cache (−80% encode calls), window/stride tuning, and AMP FP16.
• Built a 45-session, 91,800-frame thermal + RGB + depth dataset with a zero-manual-annotation labelling pipeline (YOLOv8n-Pose on RGB, projected into thermal space and thermally gated).
• Benchmarked 6 detector versions at IoU@0.5, cutting false positives on empty frames from 40.7% to 1.5%; traced dominant error to preprocessing rather than model capacity (10× FPR reduction from one line).

Project Intern (Top 20 Finalist, IoT Challenge)

FPT Software

Top 20 Finalist at the IoT Challenge 2025.

• Deployed quantized LLMs (LLaMA/Mistral with GGUF) on Raspberry Pi to build a conversational assistant.
• Engineered a full-stack solution (Frontend/Backend) for real-time shopper tracking and route optimization via BLE localization.

Research Consultant

WorldQuant

Quantitative research consultant.

• Developed predictive mathematical models (Alphas) to forecast financial instrument price movements.
• Performed statistical analysis on historical data to optimize trading signals.

Undergraduate Research Assistant

Intelligence, Computing and Communication Lab (ICCL), VinUniversity & Trinity College Dublin

Supervised by Assist. Prof. Nguyen Van Dinh.

• Investigated Open RAN architecture, 5G networks, and Task-Oriented Semantic Communication under supervision of Prof. Nguyen Van Dinh.
• Developed multi-modality AI models for smart aquaculture: computer vision for fish tracking, underwater size estimation, and image segmentation.
• Explored Diffusion Models for enhanced data generation and signal processing.
• Recipient of a $3,500 Student Research Grant to lead smart aquaculture research.

Bachelor of Science

VinUniversity

Electrical and Computer Engineering

• Signals and Systems
• Wireless Communications
• Machine Learning & Deep Learning
• Embedded Systems
• Computer Architecture

High School Diploma

Bac Ninh High School for the Gifted

Physics Major

Best Poster Award

Open Innovation Conference 2025, VinUniversity

Awarded Best Poster Award for research presented at OIC 2025.

Gold Medal

WorldQuant Challenge

Gold Medal for mathematical financial model development.

Australia-Vietnam Strategic Technologies Scholarship

Australia-Vietnam Strategic Partnership

Merit-based scholarship for strategic technology research.

Silver Medal

The University Physics Competition 2024

International university physics modeling competition.

Finalist

2025 Quantum Global Industry Challenge (Connected DMV)

Global challenge on quantum computing applications.

Finalist

IoT Challenge 2025 (FPT Software)

Top 20 Finalist for WinCart AI-powered smart shopping assistant.

Student Research Grant ($3,500)

VinUniversity

Grant lead to develop multi-sensory smart aquaculture IoT systems.

FISHER: Gradient-Decoupled Hierarchical Multi-Task Learning for Fine-Grained Recognition

Major Revision

Major Revision. Fine-grained aquatic species recognition via gradient-decoupled hierarchical multi-task learning. arXiv:2607.20523.

C-MambaPose: A Physics-Informed Complex Mamba Framework for Cross-Environment WiFi Human Pose Estimation

arXiv preprint

Physics-informed complex Mamba framework for WiFi-based human pose estimation.

ChronoSC: Task-Oriented Semantic Communication via Temporal-to-Color Encoding

2026 IEEE 11th International Conference on Communications and Electronics (ICCE 2026)

Accepted. Task-oriented semantic communication using temporal-to-color encoding.

From Monitoring to Action: A Comprehensive Survey of Autonomous Decision-Making in Smart Aquaculture Systems

Computers and Electronics in Agriculture

Major Revision. Comprehensive survey of autonomous decision-making for smart aquaculture.

Energy-Aware Resource Allocation for Energy Harvesting Powered Wireless Sensor Nodes

IEEE Communications Letters

Novel energy-aware resource allocation framework enabling asynchronous accumulate-then-transmit protocol for EH-powered WSNs.

Privacy-Preserving Thermal Fall Detection on Edge AI

Deep JSCC for Visible Light Communication

Smart Marine Aquaculture – Nha Trang

Quadcopter Design, FEA Simulation & Teensy 4.1 Flight Controller

WinCart – Edge AI Smart Shopping Assistant & Indoor Navigation

Neural Network From Scratch

1.8 GHz Microstrip Hybrid Ring (Rat-Race) Coupler