Chris Choy
Research Scientist at NVIDIA
Palo Alto, California, United States
Summary
🤩
Rockstar🎓
Top SchoolChris Choy is a research scientist at NVIDIA with 11 years of experience building and optimizing deep learning systems for 3D vision and sparse data. He holds a Ph.D. in Computer Vision from Stanford and previously worked in leading labs including Intel Labs and NEC Labs, blending rigorous research with production-focused engineering. At NVIDIA he contributes to core ML infrastructure such as the widely used MinkowskiEngine, driving low-level kernel and algorithm improvements for high-dimensional sparse tensors. His open-source work on projects like 3D-R2N2 demonstrates strong skills in data pipelines, model engineering, and system refactoring beyond novel model design. Based in Palo Alto, he combines academic depth with practical implementation, often focusing on the mathematical and computational foundations that make large-scale 3D perception systems efficient.
11 years of coding experience
Master of Science - MS, Electrica Engineering, Master of Science - MS, Electrica Engineering at Stanford UniversityBachelor of Science (BS), Electrical and Electronics Engineering, 4.12/4.3, Bachelor of Science (BS), Electrical and Electronics Engineering, 4.12/4.3 at Korea Advanced Institute of Science and Technology
English, Korean
Github Skills (18)
algorithm10
algorithms10
pytorch10
python10
sparse-matrix10
machine-learning10
recurrent-neural-networks10
deep-learning10
optimisation10
data-processing10
neural-network10
cuda10
convolution10
sparse-data10
theano10
Programming languages (5)
Github contributions (5)
chrischoy/3D-R2N2
Aug 2016 - Jul 2021
Single/multi view image(s) to voxel reconstruction using a recurrent neural network
Role in this project:
ML Engineer
ML EngineerContributions:43 commits, 4 PRs, 44 pushes in 5 years
Contributions summary:Chris primarily contributed to refactoring and restructuring the codebase related to data processing and model training within the 3D-R2N2 project. They implemented and modified data loading processes, including image preprocessing and voxel label handling, crucial for training the recurrent neural network. Furthermore, the user made significant changes to the network definition and testing procedures.
reconstructionimage-viewmulti-viewcomputer-visionrecurrent-neural-network
NVIDIA/MinkowskiEngine
Feb 2018 - Oct 2022
Minkowski Engine is an auto-diff neural network library for high-dimensional sparse tensors
Role in this project:Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:6 releases, 571 commits, 34 PRs in 4 years 9 months
Contributions summary:Chris's commits primarily focus on adding and modifying code related to the MinkowskiEngine library, which is designed for high-dimensional sparse tensors. The contributions demonstrate the user's involvement in feature additions, such as channelwise convolution and also encompass fixes for batch normalization and instance normalization. The changes also show enhancements in kernel map generation and improvements related to the core algorithms, which indicates a strong focus on the underlying mathematical and computational aspects of the library.
high-dimensional-datasemantic-segmentationsparsetensorflowhigh-dimensional
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.