Junha Lee
Research Scientist at SqueezeBits
Seoul, South Korea
Summary
🤩
Rockstar🎓
Top SchoolJunha Lee is a research scientist specializing in the intersection of efficient AI and physical AI, currently at SqueezeBits after completing a Ph.D. in Computer Science and Engineering at POSTECH. Over six years, his research evolved from 3D geometric understanding and open-vocabulary 3D vision-language models to robotic manipulation, producing projects like HighDimConv, Mosaic3D, Affogato, and DextER. He pairs strong academic grounding with practical systems work—contributing core nearest-neighbor functionality to the widely used Open3D library—demonstrating both algorithmic depth and C++ implementation skills. Junha has industry research experience at NVIDIA and a track record of shipping reproducible code and tests, making him adept at moving ideas from prototype to robust implementations. Based in Seoul, he combines a rigorous research mindset with hands-on engineering across perception, language, and embodied AI.
6 years of coding experience
박사, 컴퓨터공학, 박사, 컴퓨터공학 at Pohang University of Science and Technology
English, Korean
Stackoverflow
33reputation
2kreached
0answers
2questions
Github Skills (12)
algorithm10
data-structures10
algorithms10
c-language10
cpp10
cprogramming-language10
data-structure10
perception10
3d10
point-cloud9
machine-learning9
python6
Programming languages (3)
Github contributions (5)
isl-org/Open3D
Aug 2020 - Sep 2022
Open3D: A Modern Library for 3D Data Processing
Role in this project:
Back-end Developer
Back-end DeveloperContributions:60 reviews, 69 commits, 34 PRs in 2 years
Contributions summary:Junha contributed to the development of the "Nearest Neighbor" module, which is crucial for 3D data processing within the Open3D library. The primary focus was on implementing nearest neighbor search functionality, likely using efficient algorithms like NanoFLANN, given the code modifications. The changes include the creation and modification of C++ files, specifically `NanoFlannIndex.cpp`, which suggests the user worked on the core data structures and algorithms for searching. The user also implemented test for these functionalities as evident by changes in `NearestNeighborSearch.cpp`.
python3d-perceptionodometrytensorflowpointcloud
POSTECH-CVLab/Open3D
Apr 2020 - Jul 2020
Open3D: A Modern Library for 3D Data Processing
Contributions:19 commits, 25 pushes, 4 branches in 2 months
3d-modelspythonmodern-librarysdfdata-processing
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.