Weiyuan Kong
Computer Vision & Scientific Imaging Engineer | Applied Machine Learning
Salaire / Taux journalier
Paris, France
Freelance, ouvert au CDI
Peut venir sur site
Expertises
Langues
À propos
Introduction
Computer Vision & Scientific Imaging Engineer with a PhD in biophysics and 8+ years of experience building quantitative image-analysis pipelines, segmentation workflows, and reproducible tools for complex scientific imaging data. My background combines microscopy, computer vision, machine learning for image analysis, physical modelling, and experimental instrumentation. Previously at CNRS / Laboratoire MSC, I developed Python/MATLAB pipelines for 2D/3D microscopy analysis, deployed U-Net-based segmentation workflows across 100k+ images, and built hardware-software acquisition systems linking sensors, Arduino control and confocal imaging. I also develop open-source tools around image data quality and scientific software: → imgclean: Python CLI/API for image dataset auditing, duplicate/leakage detection and HTML/JSON/CSV reports — published on PyPI → ForceInferencePy: scientific imaging package for topology extraction and mechanical inference from microscopy images → CineDepthPro: Android computational photography prototype combining monocular depth estimation and GPU-based rendering PhD in developmental biophysics, 3 first-author publications, 80+ citations. Core skills: Python · OpenCV · scikit-image · PyTorch · TensorFlow · U-Net · C++ · MATLAB · SQL · Docker · CI/CD · FastAPI · microscopy image analysis · segmentation · tracking · quantitative imaging Open to roles in computer vision, scientific software, bioimage analysis, medtech/biotech imaging, and R&D engineering. GitHub: github.com/Weiykong
Expérience de travail
R&D Engineer – Computer Vision & Imaging Systems
CNRS – Laboratoire Matière et Systèmes Complexes
sept. 2021 - août 2025
4 ans
Paris, France
Designed and deployed computer vision pipelines and custom instrumentation for quantitative imaging of biological systems under mechanical constraints. Deployed U-Net-based 3D segmentation pipelines across 100k+ microscopy images, reducing manual analysis time by 50%. Built an Arduino/JavaScript control system synchronizing pressure-probe acquisition with confocal imaging in real time. Developed automated image processing workflows (segmentation, tracking, feature extraction) in Python for noisy, heterogeneous datasets. Created validation pipelines improving measurement robustness and reproducibility across imaging conditions. Performed multiphysics simulations (COMSOL, Abaqus) linking experimental imaging data with mechanical models.
Formation
Aix-Marseille University
Ph.D. in Developmental Biology, Quantitative mechanical measurements across scales (2D & 3D) during epithelial morphogenesis
2017 - 2021
4 ans 1 mois
Activités et associations : Interdisciplinary PhD at the interface of physics, biology and engineering. Strong involvement in experimental design, quantitative image analysis, and physical modeling.
Interdisciplinary PhD combining developmental biology, biophysics and quantitative imaging. Designed and implemented experimental setups to measure mechanical forces and growth dynamics in epithelial tissues, from cellular to tissue scales. Developed quantitative image analysis pipelines (2D/3D), combining classical image processing and applied machine learning for segmentation, tracking and feature extraction. Used physical modeling and numerical simulations to interpret experimental data and link mechanics to morphogenesis. Strong emphasis on reproducibility, automation, and cross-disciplinary collaboration.
University of Strasbourg
mster, Biophysics
2015 - 2017
1 an 11 mois
Résultat : Master
University of Strasbourg
International interdisciplinary Master, Biophysics
2015 - 2017
2 ans 1 mois
University of Strasbourg
Bachelor's Degree, Theoretical Physics
2012 - 2015
3 ans 1 mois