Marina Béguin

Marina Béguin

Internship in the Airbus Helicopters company

Followers of Marina Béguin169 followers
location of Marina BéguinZurich, Zurich, Switzerland

Connect with Marina Béguin to Send Message

Connect

Connect with Marina Béguin to Send Message

Connect
  • Timeline

  • About me

    Postdoctoral Researcher at ETH Zürich | PhD | Image Reconstruction & Computer Vision Researcher

  • Education

    • Paris Saclay University

      2017 - 2019
      PhD in Applied physics
    • Télécom Physique Strasbourg

      2013 - 2016
      Diplôme d'ingénieur généraliste spécialisé en physique et modélisation
    • Strasbourg university

      2015 - 2016
      Particle physics master degree
  • Experience

    • Airbus Helicopters

      Jun 2014 - Jul 2014
      Internship in the Airbus Helicopters company

      Observation of the Quality controller job on the Tiger Helicopter manufacturing chain

    • CERN

      Jun 2015 - Aug 2015
      Internship in the CMS collaboration

      The luminosity increase planned for the run 2 of the LHC (January 2018- December 2019) required an improvement of the CMS tracker modules. The internship mission was to took part on the production and to carry out several tests on the new assembled pixel modules.- Improvement of the manufacturing tools accuracy- Test of the new modules: production tests to check the quality of the production process and module qualification tests (thermal cycling and X-ray)- Control of the proper functioning and improvement of the test equipment Show less

    • CEA Saclay

      Mar 2016 - Aug 2016
      Internship at CERN in the FCC collaboration

      The aim of this internship was to design and optimise a detector for the FCC-ee project, a circular electron-positron collider for the post LHC physics. The measurement of the W mass at the W-pair production threshold was chosen as a benchmark, with the objective to reach a total unprecedented uncertainty of 1 MeV.

    • CERN

      Oct 2016 - Dec 2019

      Subject: Calorimetry and W mass measurement for future experimentsMy analysis addressed the calorimetry of two future experiments:- the high-granularity endcap calorimeter (HGCal) to be installed in the phase-2 upgrade of the CMS detector for the high-luminosity phase of the LHC;- the complete calorimetric system of a detector for the FCC-ee project, a future electron-positron collider in a 100 km ring, for the post-LHC physics.Using simulation, my objective was to optimise their performance in terms of particle identification and reconstruction. For the CMS part, I implemented in C++ a comprehensive, fast and flexible simulation tool that makes it possible to easily test and understand the HGCal characteristics. I was also involved in the improvement of the Geant4-based full simulation tool for the CMS detector (CMSSW) in order to optimise the HGCal jets resolution and the reconstruction of the missing energy. In a second step, the W boson was used to add constraints on the HGCal parameters and to identify the requirements to measure its mass with the best possible accuracy. In this study, I used a software using the Machine Learning Tensorflow library to improve the recoil reconstruction, key parameter in the W mass reconstruction in proton-proton collisions. The W boson mass was also used as reference for the study of a detector for the FCC-ee. With a simulation tool implemented in Python, I demonstrated that the FCC-ee current calorimetry enable the measurement of the W boson with an unprecedented statistical uncertainty. Show less

      • PhD candidate in the FCC and CMS collaborations

        Feb 2017 - Dec 2019
      • Internship in the ATLAS collaboration

        Oct 2016 - Jan 2017
    • CERN

      Mar 2020 - Jun 2020
      Fellow in the FCC collaboration

      Readout Electrode Design and Performance Optimization of for the Future Circular Collider (FCC) liquid Argon calorimeter

    • ETH Zürich

      Jul 2020 - now

      I developed a 3D image reconstruction software in C++ for the analysis of PET images, improving reconstruction accuracy without relying on external data typically used for attenuation and photon scatter corrections. This innovative approach optimizes the efficiency of analyses in a medical context.I integrated machine learning techniques, particularly Gaussian Splatting, to improve photon attenuation correction and optimize the software's performance.As a co-manager of the project, I led a clinical trial at CHUV, coordinating a multidisciplinary team, ensuring compliance with regulatory requirements, and drafting the necessary documentation for Swiss authorities.My work has been presented at several international conferences and seminars. Show less

      • Postdoctoral Researcher at the ETH Computer Vision Lab

        Jan 2025 - now
      • Postdoctoral researcher

        Jul 2020 - Dec 2024
  • Licenses & Certifications

    • Advanced Learning Algorithms

      DeepLearning.AI, Stanford University
      Mar 2024
      View certificate certificate
    • Supervised Machine Learning: Regression and Classification

      DeepLearning.AI, Stanford University
      Feb 2024
      View certificate certificate
  • Honors & Awards

    • Awarded to Marina Béguin
      Lauréate du concours Telecom Seeds for the Future Huawei France Aug 2014 Un mois en Chine à Pékin et Shenzhen comprenant deux semaines de cours intensifs de Mandarin à la BLCU et deux semaines de cours théoriques et d'application sur les futurs architectures et protocoles des réseaux (cloud computing, 4G/5G, IP) au quartier général de Huawei à Shenzhen. 12 lauréats en France.One month in China in Beijing and Shenzhen with two weeks of Chinese intensive lessons at the BLCU and two weeks of IT & telecom training at the Huawei base in Shenzhen : theoretical… Show more Un mois en Chine à Pékin et Shenzhen comprenant deux semaines de cours intensifs de Mandarin à la BLCU et deux semaines de cours théoriques et d'application sur les futurs architectures et protocoles des réseaux (cloud computing, 4G/5G, IP) au quartier général de Huawei à Shenzhen. 12 lauréats en France.One month in China in Beijing and Shenzhen with two weeks of Chinese intensive lessons at the BLCU and two weeks of IT & telecom training at the Huawei base in Shenzhen : theoretical andpractical formation on future network architectures and protocols (optical networks, cloudcomputing, 4G/5G, IP). 12 laureates in France Show less