Pierre Moulon is a Research Scientist with 16 years of hands-on experience in computer vision, photogrammetry, and large-scale 3D reconstruction, currently based in Seattle and working at Meta. He has led research and engineering efforts at Zillow to deploy production-grade 3D reconstruction and immersive home-shopping tools, and earlier helped maintain and evolve core SfM/MVS pipelines at companies like Acute3D and Auxilium. A longtime open-source contributor, Pierre has improved foundational libraries such as openMVG, VLFeat and CMVS-PMVS, focusing on robust back-end fixes, memory-leak repairs, C++11 modernization, and CI/build systems that keep research reproducible. He combines rigorous academic training in computer vision with practical software craftsmanship—optimizing performance, refactoring legacy code, and adding test suites to hard geometric algorithms. Notably, his work often targets the plumbing of spatial AI (camera intrinsics, pose estimation, export formats), making complex reconstructions reliable at scale. He’s committed to reproducible science and builds tools that let research move cleanly from prototype to production.
16 years of coding experience
9 years of employment as a software developer
Master, IT, Computer Vision and Digital Imaging, Masters of Science, Master, IT, Computer Vision and Digital Imaging, Masters of Science at Telecom Saint Etienne
Associate, IT : especially Digital Imaging and Computer Vision, Associate, IT : especially Digital Imaging and Computer Vision at UNIVERSITE D'AUVERGNE
This software (CMVS) takes the output of a structure-from-motion (SfM) software as input, then decomposes the input images into a set of image clusters of managable size. An MVS software can be used to process each cluster independently and in parallel, where the union of reconstructions from all the clusters should not miss any details that can be otherwise obtained from the whole image set. CMVS should be used in conjunction with an SfM software Bundler and an MVS software PMVS2 (PMVS version 2).
Role in this project:
Back-end Developer
Contributions:29 commits, 12 PRs, 13 pushes in 8 years 9 months
Contributions summary:Pierre implemented initial CMake-based build system for the CMVS-PMVS software, along with modifications to the `bundle.cc` file. They then incorporated the lmFit library, replacing a GSL simplex, with the aim of improving the optimization process and made several subsequent changes to the `optim.cc` and `optim.h` files, including the addition of memoization for `pow(2,X)`. Further commits focused on bug fixes and optimization of the PLY, PSET, and PATCH export functionalities.
open Multiple View Geometry library. Basis for 3D computer vision and Structure from Motion.
Role in this project:
Back-end Developer
Contributions:15 releases, 37 reviews, 1856 commits in 10 years 1 month
Contributions summary:Pierre implemented test suites for various cost functions, and also modified the build and CI setup, the camera intrinsics code, and the pose estimation algorithm. These contributions are primarily focused on the underlying mechanisms of the project, indicating a focus on the system's core components. They worked on fixing the memory leak with the EasyExif Library, improving the function signature and correcting typos.
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.