Yung-hsu Yang
PhD Candidate
Zurich, Zurich, Switzerland
Summary
👤
Senior🎓
Top SchoolYung-hsu Yang is a PhD candidate in Computer Vision at ETH Zürich with eight years of experience spanning research and applied roles in scene understanding, 3D object detection, and tracking. Trained at National Tsing Hua University and supervised by Marc Pollefeys, he blends rigorous academic research with hands-on engineering, contributing backend and data-science work to the popular Scalabel annotation tool by integrating KITTI/NuScenes formats for tracking and detection evaluation. His background includes research and QA internships in Taiwan, giving him practical exposure to real-world datasets and quality processes. Known for bridging dataset tooling and algorithmic development, he focuses on making 3D perception pipelines more interoperable and evaluable across benchmarks.
8 years of coding experience
Doctor of Philosophy - PhD, Computer Vision, Doctor of Philosophy - PhD, Computer Vision at ETH ZürichMaster's degree, Electrical and Electronics Engineering, Master's degree, Electrical and Electronics Engineering at National Tsing Hua University
Github Skills (6)
computer-vision10
data-conversion10
dataannotations10
python10
data-annotation10
scipy8
Programming languages (2)
Github contributions (5)
scalabel/scalabel
Jul 2021 - Jul 2022
Scalabel: A versatile web-based visual data annotation tool
Role in this project:
Back-end Developer & Data Scientist
Back-end Developer & Data ScientistContributions:1 review, 13 commits, 6 PRs in 1 year
Contributions summary:Yung-hsu primarily focused on converting and integrating KITTI dataset formats within the Scalabel project, a visual data annotation tool. Their contributions included adding and modifying conversion scripts, such as `from_kitti.py`, to parse and transform KITTI data into the Scalabel format, including handling of tracking and detection data types. The user also made changes to other conversion scripts, such as `to_nuscenes.py` and `from_nuscenes.py`, to improve compatibility and to add features to enable the use of the annotations for evaluation on NuScenes. These efforts involved modifications to data structures, handling of different annotation types, and adjustments to coordinate systems and rotations.
data-annotationweb-basedvisual-dataannotation-toolversatile
SysCV/qd-3dt
Mar 2021 - Jan 2023
Official implementation of Monocular Quasi-Dense 3D Object Tracking, TPAMI 2022
Contributions:1 review, 4 commits, 3 pushes in 1 year 10 months
quasicomputer-visiontracking3d-trackingobject-tracking
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.