• Name: Huynh Thai Hoa (Huỳnh Thái Hòa)
• Affiliation: Soongsil University / DeltaX
• Email: [email protected]
• Leetcode: huynhthaihoa
• Linkedin: huynhthaihoa

I am a software research engineer who takes pleasure in delving into the capability of Computer Vision and Artificial Intelligence to address everyday challenges. My goal is to contribute towards creating a better world, particularly in the areas of transportation, education, and healthcare, for the well-being of my loved ones. Throughout this lifelong and challenging journey, I eagerly anticipate collaborating with and learning from the talented individuals who surround me.

• Programming languages: C, C++, Python, C#, MATLAB, Java, Javascript.
• Fundamentals: Data Structures & Algorithms, Computer Vision, Deep Learning, Machine Learning.
• Frameworks/Libraries: OpenCV, Tensorflow, Pytorch, Darknet, FFMPEG, DeepStream, ROS, WPF, Qt.
• Software version control: Git, SVN.
• Databases: Microsoft SQL Server, MySQL, SQLite, Postgres.
• Hardware platform: Jetson Nano, Jetson Xavier NX.
• Languages: Vietnamese (Native), English (Professional), Korean (Basic).
• Others: Microsoft Office, LaTEX, JSON, XML, Batch.

• AI Researcher - DeltaX (Republic of Korea) | 06/2022 - present
  • Developing and implementing solutions for company projects about 2D/3D perception and Advanced Driver Assistant Systems, both in-cabin and outside.
  • Collaborating with the team to propose and design solutions around an Advanced Driving Assistance System using cameras (pinhole/fisheye, RGB/IR) with LIDAR supervision.
  • Designing, implementing, and maintaining utility software for camera calibration, data collection, annotation, and analysis.
• Research Engineer - Smart Deep (Republic of Korea) | 05/2021 - 05/2022
  • Participated in Deep Neural Network training pipeline: data labeling, cleaning, and training.
  • Developed tools for preprocessing datasets and evaluating inference model consistency.
  • Developed new features and improved client programs’ performance and dynamic-link libraries (DLL).
  • Developed a video processing system for interesting object detection.
  • Researched and proposed solutions for company projects.
• Research Intern - Ubay Solution (Republic of Korea) | 12/2020 - 04/2021
  • Researched to find solutions for Optical Character Recognition (OCR) problems in detecting and recognizing license plate numbers and container codes.
  • Evaluated the performance and consistency of several deep learning frameworks on our custom dataset.
  • Deployed inference model on NVIDIA Jetson Nano board
• Research Assistant - ANDA Lab (Republic of Korea) | 08/2018 - 08/2020
  • Researched to find and solve problems in the vehicle localization process based on detecting vehicle taillights using camera images.
  • Read and wrote scientific papers related to the project.
• Trainee - Global Cybersoft JSC (Vietnam) | 03/2017 - 08/2017
  • Learned the software development process in practice.
  • Supported team to design and develop plugins as required by customers.
  • Attended the peer-review process that uses appropriate tools to evaluate code quality.

• Master of Engineering in Information Communication - Soongsil University (Republic of Korea) | 08/2018 - 08/2020
  • GPA: 4.01/4.05
  • Thesis: LED Taillight Matching in Nighttime Vehicle Positioning
  • Supervisor: Professor Myungsik Yoo
• Bachelor of Engineering in Software Engineering - University of Information Technology, Vietnam National University in Ho Chi Minh City (Vietnam) | 08/2013 - 06/2018
  • GPA: 8.6/10
  • Thesis: Coronary Arteries Segmentation from Coronary Angiogram
  • Supervisor: M.Sc. Thi-Vuong Pham

• Thai-Hoa Huynh, Myungsik Yoo, Taillight Matching and Pairing Algorithm for Stereo-Vision-based Nighttime Vehicle Positioning, Applied Sciences (2020)
• Thai-Hoa Huynh, Myungsik Yoo, Nighttime LED Taillight Detection Method with Optical Camera Communication System, 한국통신학회논문지 (2020)
• Thai-Hoa Huynh, Tuan-Anh Pham, Myungsik Yoo, Detection Algorithm for Overlapping LEDs in Vehicular Visible Light Communication System, IEEE Access (2019)