_
M.S. graduate in Computer Science from Chang Gung University (GPA 3.85/4.0, Full Scholarship) with two years of AI research experience at WINSLAB. Specializing in computer vision and multimodal learning for medical imaging, including a diagnostic system achieving 90% AUC for Liver Fibrosis classification. Interested in large language models for medical imaging and vision-language model research.
Technical Skills
Deep Learning & AI
PyTorch
PyTorch Lightning
HuggingFace Transformers
ONNX Runtime
OpenCV
Medical Image Analysis
Multi-modal Fusion
VLM & Computer Vision
MLOps & Cloud
Docker
DVC (Data Version Control)
AWS (S3, ECR, Lambda)
GitHub Actions CI/CD
Hydra Configs
Web & Software Development
Magento 2
FastAPI
React
Streamlit
jQuery
RESTful APIs
SQLite
Git & Linux
Programming Languages
Python
C/C++
SQL
PHP
JavaScript
HTML/CSS
Research & Languages
Scientific Paper Writing
Literature Review & Design
English (Proficient)
Vietnamese (Native)
Chinese (Beginner)
Honors & Awards
- 2025: 🏆 1st Place, Taiwan AI Safety Challenge (TAISC) @ IEEE AVSS 2025.
- 2025: 🏅 Top 5 Finalist, Micro-Action Challenge (MAC) @ 2025 Workshop.
- 2024: 🎓 Full Scholarship (Tuition & Stipend), Chang Gung University.
- 2023: 🌟 Best Project Employee of the Quarter, FPT Software - Recognized for technical tenacity and problem-solving excellence.
- 2020: 🥉 Top 3, Hackathon at UIT.
Projects
Publications
-
VL-FOG: A Vision-Language Flow Gating Framework for Early Accident Anticipation
Anh-Khang Nguyen, et al. — IEEE AVSS Workshops, 2025
-
A Lightweight Multi-Granularity Framework with Frame-Level and Video-Level Heads for Micro-Action Recognition
Anh-Khang Nguyen, et al. — IEEE Transactions on Multimedia (TMM), 2026 (Under Review)
Experience
Research Assistant
WINSLAB, Chang Gung University | Feb. 2024 - Jan.
2026
- Independently developed a multimodal diagnostic AI system combining Nakagami parametric imaging with clinical patient data, utilizing a ResNet-18 backbone to achieve a 90% AUC for Liver Fibrosis stage classification.
- Researched and developed a Vision-Language Model (VLM) workflow for early accident anticipation, designing evaluation benchmarks and iterating through experimental analysis to outperform state-of-the-art baselines; awarded 1st Place at the Taiwan AI Safety Challenge (IEEE AVSS 2025).
- Designed a lightweight multi-granularity deep learning framework for micro-action recognition integrating frame-level and video-level heads, delivering 24x higher throughput than state-of-the-art methods — submitted to IEEE Transactions on Multimedia (under review).
Software Engineer
FPT Software | Nov. 2021 - Feb. 2024
- Maintained and scaled the official e-commerce platform for Canon Canada (shop.canon.ca) built on the Magento 2 framework using PHP, JavaScript, jQuery, and SQL, integrating third-party APIs and optimizing core legacy modules to improve performance.
- Recognized as Best Project Employee of the Quarter (2023).
Front-end Developer
cisbox Software Ltd | Sept. 2021 - Nov. 2021
- Developed analytical dashboards for a factory management platform, visualizing complex industrial data using React and TypeScript.
Education
M.S. in Computer Science and Information Engineering
Chang Gung University | Feb. 2024 - Jan. 2026
- GPA: 3.85/4.0 | Full Scholarship Recipient
- Thesis: A Lightweight Multi-Granularity Framework with Frame-Level and Video-Level Heads for Micro-Action Recognition
- Advisor: Prof. Jenhui Chen, Dept. of CSIE, CGU
B.S. in Information Technology
University of Information Technology (UIT) | Sep. 2018
- June 2022
- GPA: 3.0/4.0