Summary
Haoyuan Zhang is a senior systems software engineer focused on computer vision, visual SLAM, and AR algorithms, with eight years of experience building mapping, localization, and sensor-fusion solutions for robotics and MR platforms. Currently at NVIDIA SW-TEGRA, he applies C++, Python, and CUDA to production-grade mapping and localization stacks after leading MR spatial anchor and VSLAM work at ByteDance/PICO. His background spans academia (UPenn GRASP Lab, MSE in Robotics) and industry (MEGVII, Horizon, DJI), where he delivered HD mapping, ADAS perception, and fast loop-closure systems for real robots. He combines rigorous geometric SLAM expertise (SFM, bundle adjustment, PnP) with practical engineering to close the gap between research and deployable systems. Notably, his trajectory includes both teaching core computer vision courses and shipping GPU-accelerated perception modules, showing a rare mix of pedagogy, research, and production engineering. Based in Shanghai, he brings deep hands-on experience integrating algorithms into embedded and large-scale platforms.
8 years of coding experience
5 years of employment as a software developer
Bachelor of Engineering (BE), Computer Science, 3.75/4.00, Bachelor of Engineering (BE), Computer Science, 3.75/4.00 at The Hong Kong Polytechnic University
Bachelor of Engineering (B.E.), Electrical and Electronics Engineering, 3.90/4.00, Bachelor of Engineering (B.E.), Electrical and Electronics Engineering, 3.90/4.00 at Beijing Institute of Technology
Master of Science (M.S.), Mechatronics, Robotics, and Automation Engineering, 3.83/4.00, Master of Science (M.S.), Mechatronics, Robotics, and Automation Engineering, 3.83/4.00 at University of Pennsylvania
English, Chinese