Qingyong Hu
Ph.D. Candidate
Oxford, England, United Kingdom
Summary
🤩
Rockstar🎓
Top SchoolQingyong Hu is a DPhil candidate in Computer Science at the University of Oxford with eight years of experience focused on 3D computer vision and robotics, specializing in point cloud semantic and instance segmentation and local surface matching. He combines strong research supervision from leading academics with hands-on engineering, exemplified by his TensorFlow implementation and enhancements of the influential RandLA-Net model (CVPR 2020 / IEEE TPAMI 2021). His work spans developing core network architectures, training/evaluation pipelines, and practical applications for urban-scale 3D understanding, evidenced by participation in The Alan Turing Institute’s SenSat study. Based in Oxford, he brings both academic rigor and applied ML engineering to problems enabling safer autonomy and accurate digital twins.
8 years of coding experience
Dphil in Computer Science, Computer Science, Dphil in Computer Science, Computer Science at University of Oxford
madarin chinese, English
Github Skills (5)
semantic-segmentation10
computer-vision10
tensorflow10
python10
3d10
Programming languages (3)
Github contributions (5)
QingyongHu/RandLA-Net
Nov 2019 - Jul 2021
🔥RandLA-Net in Tensorflow (CVPR 2020, Oral & IEEE TPAMI 2021)
Role in this project:
ML Engineer
ML EngineerContributions:47 commits, 1 PR, 46 pushes in 1 year 7 months
Contributions summary:Qingyong appears to be focused on the implementation and modification of a RandLA-Net model for point cloud semantic segmentation. The commits reveal changes to the core network architecture, including encoder and decoder blocks, and loss calculations. The user is also involved in setting up the training and evaluation pipelines, with modifications to the main scripts and testing procedures.
s3dissemantic-segmentationsemantic3ddeep-learningcvpr-2020
QingyongHu/SoTA-Point-Cloud
Dec 2019 - Jun 2021
🔥Deep Learning for 3D Point Clouds (IEEE TPAMI, 2020)
Contributions:45 commits, 42 pushes, 2 branches in 1 year 5 months
kitti-datasetpoint-cloudsstereodeep-learningdescriptor
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.