Zexi Han is a Staff R&D Engineer with a decade of experience building end-to-end perception and big-model algorithms for autonomous driving, currently focusing on camera object detection at Momenta. He has led dynamic LiDAR perception teams and shipped production-grade 3D detection and tracking systems that combine PointRCNN/PV-RCNN variants, optical flow attention, and real-time C++ preprocessing optimizations. Zexi’s background spans applied research and engineering across startups and research institutes, from creating large-scale annotated 3D datasets and bespoke labeling tools to deploying model pipelines in constrained embedded environments. Based in Shanghai, he blends hands-on deep learning, sensor fusion, and systems engineering with a track record of improving both accuracy and latency in safety-critical perception stacks. An interesting detail: he has repeatedly moved models from SOTA research into real-time product settings, achieving metrics like 0.93 AP on pedestrian detection through iterative annotation and algorithmic tuning.
10 years of coding experience
5 years of employment as a software developer
BS, Telecommunications Engineering, BS, Telecommunications Engineering at Beijing University of Posts and Telecommunications
MS, Data Science, MS, Data Science at Northeastern University
BS, Telecommunications Engineering, BS, Telecommunications Engineering at Queen Mary University of London
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.