Matthieu Zins

Matthieu Zins

Computer Vision engineer (PhD)

Inria Nancy - Grand Est

About me

I am a PhD student in the Tangram team (previously Magrit) at Inria in Nancy. I am supervised by Marie-Odile Berger and Gilles Simon. My general areas of interest are computer vision, machine learning and augmented reality. My research focuses on improving visual localization using abstract object models.

Before starting my PhD, I have worked for 2 years at Kitware, as a computer vision engineer. I wrote my Master Thesis at SICK IVP, supervised by Ola Friman, on Super-resolution for Time-of-flight Cameras. I received a French Engineering degree from the UTC and a MSc from Linköping University, Sweden, both in Computer Science. I have also spent 6 months as an exchange student at Technische Universität Chemnitz in Germany.

News

Interests

  • Computer Vision
  • Augmented Reality
  • Deep Learning
  • Algorithms

Education

  • PhD in Computer Vision, 2022

    Inria

  • MSc in Computer Science, 2017

    Linköping University

  • Engineer degree in Computer Science, 2017

    Université de Technologie de Compiègne

Experience

 
 
 
 
 

PhD Student in Computer Vision / Augmented Reality

Inria

Oct 2019 – Sep 2022 Nancy, France
The goal of my thesis is to improve Visual Localization accuracy and robustness in complex environments using high-level object landmarks. In practice, we combine geometrical reasoning with recent Deep Learning approaches for object detection and pose estimation.

  • Development of an object-based visual SLAM system offering automatic object mapping and robust relocalization.
  • Publications in international journal and conferences: IJCV, 3DV, IROS, ISMAR.
  • Codes (Python and C++) released at gitlab.inria.fr/tangram.
  • Manuscript | Examinators report
 
 
 
 
 

Computer Vision Engineer

Kitware

Oct 2017 – Sep 2019 Lyon, France
I have worked on different projects involving 3D Reconstruction, SLAM, Sensor Calibration, Point Cloud Analysis, Texture Mapping and Satellite Imagery.

  • Algorithms development for different sensors: Kinect Azure, Pico Flexx, Intel RealSense depth and tracking cameras.
  • Contributions to KWIVER, an open-source toolkit for Computer Vision (C++).
  • Development of texture mapping algorithms for a large research project about urban semantic 3D reconstruction from multi-view satellite imagery. This work received the Best Paper Award at CVPR Workshop 2019.
  • Scientific papers review and presentation to the team.
 
 
 
 
 

Master Thesis

SICK IVP

Jan 2017 – Oct 2017 Linköping, Sweden
I wrote my Master thesis at SICK IVP, in collaboration with the Computer Vision Laboratory (CVL) at Linköping University. I have worked on:

  • Sensor fusion between a time-of-flight camera and a color camera.
  • Super-resolution techniques for depth cameras.
 
 
 
 
 

Software Engineer Intern

DeltaCAD

Aug 2015 – Feb 2016 La Croix-Saint-Ouen, France
Algorithmic processing for a Virtual Reality application (C++):

  • Parallelization of geometric the processings with multithreading.
  • Recognition of 3D annotations.
  • Optimization of the import of 3D models: obj, 3dxml, collada, vrml, stl.
  • Optimization of the 3D viewer.
  • Automatic deployment in a VR environment.

Competitions

Data Science & Programming competitions