Postdoctoral Position (M/F) in Chemical Characterization of Pictorial Layers Using AI and Two- and Four-Flux Models

Job Description

Missions

The postdoctoral researcher will develop an approach for the analysis of hyperspectral imaging data applied to pictorial layers, based on coupling physical radiative transfer models (two-flux and four-flux approaches) with machine learning methods. The researcher will implement and train neural network architectures, including Physics-Informed Neural Networks (PINNs), in order to integrate physical constraints into the learning process and improve the identification and quantification of pigments and binders. The work will include the generation of simulated spectral data based on physical models, the training and optimization of machine learning models, and their validation on hyperspectral imaging data acquired from cultural heritage objects.

Activity

The postdoctoral researcher will conduct research aimed at developing and implementing hybrid models that combine physical radiative transfer modeling with deep learning for the analysis of hyperspectral imaging data. The researcher will be responsible for the design and development of numerical models, including neural network architectures that integrate physical constraints (Physics-Informed Neural Networks), as well as for the implementation and optimization of the associated algorithms. The researcher will analyze and interpret experimental and simulated spectral data, evaluate the performance of the developed models, and contribute to their methodological improvement. They will also contribute to scientific exchanges within the laboratory and with project partners, and to the dissemination of results through scientific publications, presentations, and the sharing of developed data and tools.

Your Profil

Skills

The candidate must hold a PhD in physics, chemistry, optics, signal processing, data science, or a related discipline. They should have strong skills in physical modeling and/or machine learning, as well as experience in the development of scientific algorithms (Python, Matlab, or equivalent). Experience with deep neural networks, particularly for the analysis of spectral or imaging data, will be highly appreciated. Knowledge of light propagation in scattering media, spectroscopy, or hyperspectral imaging would be an asset. The candidate should also demonstrate autonomy, scientific rigor, and the ability to work in an interdisciplinary environment. A good command of scientific English (both written and spoken) is required for writing publications and communicating results.

Your Work Environment

The postdoctoral researcher will be hosted at the Centre de Recherche sur la Conservation (CRC – UAR 3224 CNRS / MNHN / French Ministry of Culture), within the Colours and Visual Effects research group. The project is part of the field of heritage science and aims to develop innovative methods for the non-invasive analysis of pictorial materials using hyperspectral imaging data. The researcher will work in an interdisciplinary environment at the interface of physics, chemistry, data science, and the conservation and restoration of cultural heritage. They will benefit from the laboratory's expertise in the characterization of pictorial materials, optical modeling, and scientific imaging, as well as access to hyperspectral imaging equipment and to heritage collections studied within the laboratory's activities.

Compensation and benefits

Compensation

33700-38700

Annual leave and RTT

44 jours

Remote Working practice and compensation

Pratique et indemnisation du TT

Transport

Prise en charge à 75% du coût et forfait mobilité durable jusqu’à 300€

About the offer

About the CNRS

The CNRS is a major player in fundamental research on a global scale. The CNRS is the only French organization active in all scientific fields. Its unique position as a multi-specialist allows it to bring together different disciplines to address the most important challenges of the contemporary world, in connection with the actors of change.

CNRS

The research professions