CNRS Internship – Photoluminescence Imaging: Analysis of Big Data by Coupling Physical Models and Neural Networks


Function: Master student

Contract Type: Internship

Starting Date: April 2022

Working Place: Palaiseau, France (Paris-Saclay technology cluster)

Duration: 6 months

Education: Master 2




Become an actor of the Energy Transition by joining a team driven by innovation and impact to address today’s most decisive challenges.


IPVF – Institut Photovoltaïque d’Île-de-France, is a global Research, Innovation and Education center, which mission is to accelerate energy transition through science & technology.
Gathering industrial PV leaders (EDF, TotalEnergies, Air Liquide, Horiba and Riber) and world-renowned academic research organizations (CNRS, Ecole Polytechnique), multi-disciplinary and international IPVF teams conduct research for clean energy technologies. Supported by the French State, IPVF is labelled Institute for Energy Transition (ITE).


IPVF at a glance:
• An ambitious Scientific and Technological Program (6 programs divided in 24 work packages): from tandem solar cell technologies to economy & market assessment, state-of-the art characterization, photocatalysis and breakthrough concepts.
• State-of-the-art technological platform (8,000m²): more than 100 cutting-edge equipments worth €30M, located in cleanrooms (advanced characterization, materials deposition, prototypes for fabrication, modelling…).
• High-standard Education program (M.S. and PhD students).



University Paris-Saclay (UPS) has more than 30,000 students, nearly 20% of all French research, and several of the most innovative companies. It is not only the number one European scientific hub, but also figures among the top 20 university clusters in the world. CentraleSupélec (CS) is a major player in the development of UPS. It offers a broad scientific spectrum, covering all Engineering and Systems Sciences. Research activities in CS are developed in various research domains: Applied Mathematics, Signal processing and Automatic Control, Materials-Procedures, Mechanics-Energetics-Combustion, Applied Physics, Technology and IT Systems, Electrical Engineering and Electronics, Industrial Engineering-Economics-Administration. CS is associated with the main national research organizations (CNRS, Inria, CEA, etc.).



The project will be carried out at IPVF and CentraleSupélec, both located in the heart of the new Paris-Saclay campus.



Photovoltaic (PV) devices offer a direct conversion of light source into electricity, providing a much-needed solution to meet climate targets and move towards a low-carbon economy. In the solar cell community there is a growing interest in the use and in the optimisation of optical imaging techniques, mainly based on photoluminescence (PL) analysis. IPVF optical characterisation lab is at the forefront of the development of advanced characterisation methods for solar cells with a solid expertise on hyperspectral luminescence and Time-Resolved FLuorescence Imaging (TR-FLIM). With these solely optical technique –therefore applicable at any stages of a solar cell fabrication process- it is possible to image in a quantitative way a large panel of optoelectronics properties such as the potential voltage, defects density, diffusion lengths, absorption properties. To further expand the capabilities of hyperspectral PL datasets, it is possible to process the signal by using different algorithms to improve the signal-to-noise ratio and to identify possible correlations between different features within the material. The implementation of this approach is expected to provide new insights and a more detailed and comprehensive knowledge about the local optoelectronics properties of photovoltaic materials.



This 6-months internship position, carried out in collaboration between the IPVF-UMR and the Centre for Visual Computing of CentraleSupélec & Inria, focuses on the development of advanced variational techniques to process complex 3-dimensional PL datasets. Specifically, the candidate will use parallel proximal optimization algorithms allowing to introduce suitable spatial regularisation on the sought data while taking into account the existing physical priors. In addition, the proposed algorithms will be unrolled under the form of deep neural networks to allow faster computations and automatic tunning of hyperparameters.




  • Strong background in mathematics, image processing, and machine learning
  • Matlab, Python programming skills




CV and cover letter to be sent to :

Need a direct line?

Feel free to contact us for more information about our offers.

  • +33(0)1 69 86 58 60
Contact us