Technical Skills

Languages: Python, C/C++, Matlab, Labview, Maple

Hardware: Raspberry Pi, Jetson Nano, Arduino

Developer Tools: Git, Docker, Google Cloud Platform, VS Code, Visual Studio, Jupyter Notebook

Libraries: Pandas, NumPy, Matplotlib

Education

Sep 2021 - Mar 2024: University of Trento, Trento, Italy

Master in Electronics and Robotics
Sep 2013 - Sep 2018: Hanoi University of Science and Technology, Hanoi, Vietnam

Degree of Engineer in Mechatronics Engineering

July 2024 - Now: Robotics Research Fellow

Istituto Italiano di Tecnologia, Genoa, Italy

• Research and develop a lightweight algorithm for 3D scene estimation and human pose estimation using event camera.
Jun 2023 - Mar 2024: Computer Vision Research Intern

Viscando AB, Gothenburg, Sweden

• Thesis project: 3D object detection and tracking based on Deep Neural Networks using stereo camera sensors.

• Utilizing stereo camera sensors developed by Viscando to extract 3D bounding boxes of objects, contributing to enhancing 3D & AI-based sensor technology for safety on public roads and in intralogistics.
Jan 2020 - Oct 2022: Research Assistant

Phenikaa University, Hanoi, Vietnam

• Implemented Deep Learning models on Ultra-wideband radar for people counting applications with 97% accuracy running on NVIDIA Jetson Nano.

• Tutored 30 students for the Computer Vision course.

Mar 2023 - Jun 2023: Mow-E (collaborated with Husqvarna Robotics)

C++, Python, Git

• Implemented Particle Filtering for SLAM in Husqvarna Robot Mower, leveraging LiDAR sensor data to ensure precise navigation and mapping capabilities.

• Collaborated with the back-end team to establish a robust REST API infrastructure, enabling real-time tracking, monitoring, and control of the mower’s position and operations.
Mar 2023 - Jun 2023: SINTIA

Python, Pandas, TensorFlow, Git

• Developed an abnormal pedestrian behavior detection system based on their trajectories, enhancing the perception in stereo cameras developed by Viscando AB.

• Pre-processed and analyzed time series data extracted from the stereo cameras.
July 2024 - now: Human Pose Estimation

Python, C++, Pytorch Lightning, Git, Docker

• Develop a lightweight algorithm to estimate human pose that can be run at 100 Hz using event camera and graph neural networks.

• Write a Dockerfile to containerize a real-time human pose estimation algorithm, ensuring compatibility of dependencies, system stability, and enhanced deployment efficiency for seamless execution in Docker environments.