Data scientist / Computer vision on ancient manuscripts (M/F)

Job Description

Missions

The position holder adapts and refines the Vesuvius Challenge's open-source pipelines (surface segmentation, virtual unrolling, ink detection using neural networks) to apply them to tomography data from other scrolls published in June 2026. Within the ERC PALAI project, he or she will also contribute to the development of a computer vision methodology inspired by the Vesuvius Challenge tools, applicable to the entire corpus of palimpsest manuscripts studied by the project.

Activity

Adaptation of Vesuvius Challenge Tools:
- Adapt the Vesuvius Challenge's open-source pipelines (surface segmentation, virtual unrolling, ink detection using neural networks) to apply them to recently published charred scroll tomography data;
- Preprocess and register the processed tomography volumes, then train, validate, and refine ink detection models adapted to the scarcity of annotated data.

Methodological Development for the PALAI Corpus:
- Design and formalize, building on this adaptation work, a generic computer vision methodology applicable to the entire corpus of palimpsest manuscripts studied by the project;
- Test ink detection models adapted to the scarcity of annotated data.

Knowledge Transfer:
- Document the code and methodology developed, with a view to their reuse by the team on other manuscripts in the corpus;
- Contribute to the scientific dissemination of results derived from this methodology.

Your Profil

Skills

Knowledge:
- Computer vision and deep learning (image segmentation, convolutional networks, generative methods)
- 3D volume processing / scientific or medical imaging
- Familiarity with Vesuvius Challenge tools and literature desired
- Awareness of conservation issues and the material diversity of palimpsest manuscripts (supports, states of conservation, acquisition methods), useful for methodological generalization

Operational skills:
- Proficiency in Python and standard scientific libraries (PyTorch or TensorFlow, NumPy, scikit-image)
- Ability to get up to speed on an existing scientific codebase, adapt it, and extend it
- Experience with learning from limited annotated data (semi-supervised learning, synthetic data, transfer learning)
- Ability to formalize a generic methodology from applied development work, and to document it for reuse by other team members

Behavioral skills:
- Autonomy and ability to document one's work for reuse by the team
- Aptitude for working in an interdisciplinary context (philology, imaging, AI)
- Fluent scientific English

Your Work Environment

The Centre de recherches sur la pensée antique, better known as the Centre Léon Robin, is a joint research unit (UMR 8061) under the dual supervision of Sorbonne Université and the CNRS. It brings together, for the most part, CNRS researchers and Sorbonne Université faculty members. The Centre, currently directed by Marwan Rashed (Sorbonne Université) with Alexandra Michalewski (CNRS) as deputy director, is tasked with fostering and coordinating research on ancient thought and its history. Members of the Centre examine texts for their philosophical content, while giving full weight to philological and historiographical questions as well as to the political, religious, and aesthetic background of these texts. The Centre places significant emphasis on the history of their reception and on the stakes involved in their past and contemporary uses.

The successful candidate will work under the direction of Victor Gysembergh, head of the ERC Consolidator Grant PALAI project. This project, funded by the European Research Council over a period of 60 months, aims to develop and apply multispectral and synchrotron imaging methods for the non-invasive reading of palimpsest and damaged manuscripts, at the intersection of classical philology, scientific imaging, and artificial intelligence. The project runs imaging campaigns with major European libraries and heritage institutions, and has already led to several major discoveries that received international media coverage. It thus offers a particularly stimulating working environment, at the intersection of fundamental research in ancient studies and methodological innovation in computer vision, within a small interdisciplinary team (engineers, doctoral students, postdoctoral researchers).

Compensation and benefits

Compensation

between €2,991.58 and €4,756.76, depending on experience

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