Fabien Servant
Ingénieur Recherche Développement at Technicolor Group
Brittany, France
Summary
🤩
Rockstar🎓
Top SchoolFabien Servant is a computer vision R&D engineer with 8+ years of industry experience specializing in camera kinematics estimation for robotics, vehicle behavior and augmented reality, now applying that expertise to cinema post-production at Technicolor. He combines a PhD-trained scientific approach with hands-on skills in embedded optimization, delivering end-to-end solutions from algorithm design to production-ready hardware and software (notably IMU/UWB and IMU+CV camera tracking systems). Fabien has driven practical advances in ultra-large HDR panorama pipelines and contributed backend improvements to prominent open-source projects like AliceVision/Meshroom, enhancing panorama and photogrammetry workflows. Comfortable across heterogeneous platforms—from tiny ARM devices to cloud and GPGPU—he excels at squeezing computer vision into constrained environments while keeping an eye on real-world deployment constraints.
7 years of coding experience
9 years of employment as a software developer
Doctor of Philosophy (PhD), Compute science, Doctor of Philosophy (PhD), Compute science at Université Rennes 1Research master, Complex Systems modelization and simulation, Research master, Complex Systems modelization and simulation at ULCO
French, English
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (13)
computer-vision10
c-language10
eigen10
cprogramming-language10
python10
photogrammetry10
3d-reconstruction10
workflow-automation8
algorithms8
data-structures8
algorithm8
data-structure8
computer-engineering8
Programming languages (5)
Github contributions (5)
alicevision/AliceVision
May 2018 - Jan 2023
Photogrammetric Computer Vision Framework
Role in this project:
Back-end & Computer Vision Engineer
Back-end & Computer Vision EngineerContributions:83 reviews, 665 commits, 209 PRs in 4 years 8 months
Contributions summary:Fabien primarily contributed to the core functionality of the photogrammetric computer vision framework, specifically modifying the pretty print feature to avoid Eigen types. Their work involved fixing errors in the equirectangular to spherical coordinate transformation and adjusting code that interacted with static arrays on different compilers. Additionally, the user removed landmarks from panorama estimation and added the constraint2D concept to the data. This involved in-place editing code.
structure-from-motionvision-frameworkpythonphotogrammetryvision
alicevision/Meshroom
Oct 2019 - Jan 2023
3D Reconstruction Software
Role in this project:Backend Developer
Backend DeveloperContributions:48 reviews, 97 commits, 100 PRs in 3 years 4 months
Contributions summary:Fabien primarily focused on enhancing the functionality of the Meshroom software by implementing new nodes and features related to panorama processing. Their contributions include adding nodes for rotating head XML, controlling panorama width, integrating LDR to HDR conversion, downscaling cameras, and splitting compositing and warping processes. The user also added options for multiband processing, HDR bypassing, and rotation offsets, demonstrating a focus on improving panorama creation workflows.
structure-from-motionreconstructionphotogrammetryhdr-imagingcomputer-vision
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.