Summary
Zhengdao Li is a computer vision analyst and MSc candidate in Computer Science based in Hong Kong with 11 years of hands-on experience applying machine learning to real-world systems. He has bridged research and production across academia and industry, from fine-tuning ViTs for hairstyle classification and facial mesh/pose pipelines at HKU to deploying perception stacks for railway safety using YOLOv8, OpenPCDet and DeepStream at Kodifly. Comfortable across the full ML lifecycle, he combines Python-first model development, MLOps (Docker, Ansible, Jenkins), and embedded/robotic integration (ROS, Fast-LIO) to deliver end-to-end solutions. His work uniquely spans Web/Unity client integrations for avatar animation and browser-based Mediapipe/three.js visualizations as well as sensor fusion for LiDAR-camera calibration. A former senior big data engineer turned PhD-track researcher, he pairs production engineering rigor with a researcher's eye for novel model fusion and efficient deployment.
11 years of coding experience
2 years of employment as a software developer
Exchange Student, Computer Science, A, Exchange Student, Computer Science, A at McGill University
The University of Hong Kong (HKU)
undergraduate, Information Technology, First Class Honor, undergraduate, Information Technology, First Class Honor at City University of Hong Kong