Jiawen Chen
Principal Researcher at Waabi
San Jose, California, United States
Summary
🤩
Rockstar🎓
Top SchoolJiawen Chen is a Principal Researcher based in San Jose with 11 years of experience at the intersection of computer vision, image processing, and high-performance systems, holding a PhD from MIT. He has moved between top-tier research and product teams at Google, Adobe, and now Waabi, advancing real-time imaging and ML-driven enhancement pipelines. His open-source contributions include performance-sensitive work on Halide and HDRNet—adding Android camera integrations and optimizing CUDA/JAX kernels for numerical accuracy and speed. Colleagues rely on him for bridging low-level C++/CUDA performance engineering with Java/Android application layers, and for pragmatic research that ships. An uncommon strength is his track record of improving both algorithmic interpolation accuracy and platform-specific lifecycle robustness, which helps turn research prototypes into production-capable systems.
12 years of coding experience
15 years of employment as a software developer
PhD, Electrical Engineering and Computer Science, PhD, Electrical Engineering and Computer Science at Massachusetts Institute of TechnologyStuyvesant High School
Stackoverflow
239reputation
21kreached
7answers
3questions
Github Skills (21)
c-language10
image-processing10
halide10
machine-learning10
jna10
android10
cuda10
computer-vision10
jax10
cprogramming-language10
mobile-development9
numerical-methods9
mobile9
tensorflow8
android-ndk6
Programming languages (16)
Github contributions (5)
google/hdrnet
Aug 2017 - Sep 2020
An implementation of 'Deep Bilateral Learning for Real-Time Image Enhancement', SIGGRAPH 2017
Role in this project:
ML Engineer
ML EngineerContributions:7 commits, 6 pushes, 3 comments in 3 years 2 months
Contributions summary:Jiawen migrated and modified JAX implementations of the `bilateral_slice` function and its application to the `hdrnet` repository. They refactored code related to trilinear interpolation, range weights, and spatial weights, while also transitioning the code to use `nda::array_ref` and updating the dependencies. Furthermore, they modified the CUDA kernels for `bilateral_slice_apply` and `bilateral_slice` to include half-pixel offsets and optimized interpolation weights, showcasing a focus on performance and numerical accuracy. This indicates a strong involvement in the core computational aspects of the HDRnet project.
image-enhancementsiggraph
halide/Halide
May 2015 - Apr 2021
a language for fast, portable data-parallel computation
Role in this project:Mobile Developer (Android)
Mobile Developer (Android)Contributions:11 reviews, 9 commits, 14 PRs in 6 years
Contributions summary:Jiawen primarily contributed to the Android-specific application code, focusing on camera functionality within the Halide framework. Their work included adding a new Android camera2 and Halide edge detector example, improving the SurfaceView lifecycle integration, and fixing stride-related issues. They also added and refined JNI and Java buffer wrappers, demonstrating proficiency in both Java and C++ within the Android environment for image processing tasks.
data-parallelhalidehexagoncompilerdsl
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.