Hongyuan Du
Senior Machine Learning Engineer, Perception at Cruise
San Carlos, California, United States
Summary
👤
Senior🎓
Top SchoolHongyuan Du is a Senior Machine Learning Engineer specializing in perception with eight years of experience building and researching deep learning solutions for object detection, segmentation, pose estimation, and self-supervised pretraining. Based in San Carlos, he currently contributes to Cruise’s Computer Vision Foundation team and previously led research efforts at TuSimple to develop keypoint-based pose estimation and scalable self-supervised image encoders for diverse, unlabeled driving data. His academic work at UC San Diego produced an ECCV-accepted proposal module for point-cloud detection and explored radar-aided monocular 3D detection, reflecting a strong bridge between novel research and production-ready systems. Hongyuan also contributes practical tooling—such as improving KITTI LiDAR visualization—demonstrating attention to reproducibility and developer ergonomics beyond model design.
8 years of coding experience
5 years of employment as a software developer
Junior and senior high school, Junior and senior high school at The Experimental High School Attached to Beijing Normal UniversityUniversity of California, San DiegoBachelor's degree, Electronic Engineering, Bachelor's degree, Electronic Engineering at Tsinghua University
English, Chinese
Github Skills (6)
3d-object-detection10
visualization10
visualizations10
python10
mayavi10
jupyter-notebook9
Programming languages (3)
Github contributions (5)
kuixu/kitti_object_vis
Jun 2020 - Jun 2020
KITTI Object Visualization (Birdview, Volumetric LiDar point cloud )
Role in this project:
ML Engineer
ML EngineerContributions:15 commits, 1 PR in 3 days
Contributions summary:Hongyuan focused on setting up and testing the visualization tools for the KITTI object detection dataset, as suggested by the repository description and topics. Their contributions primarily involve integrating the Mayavi library for 3D visualization, simplifying local installation, and removing unused variables. They also addressed minor code issues to ensure consistency.
kitti-datasetkittipoint-cloudslidar-point-cloud3d-object-detection
dashidhy/SMOKE
Jun 2020 - Sep 2020
SMOKE: Single-Stage Monocular 3D Object Detection via Keypoint Estimation
Contributions:36 commits, 36 pushes, 1 branch in 3 months
kitti-datasetdeep-learningobject-detectioncomputer-vision3d-object-detection
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.