Informations générales
Intitulé de l'offre : M/F Digital humanities research engineer, responsible for corpus creation and processing (H/F)
Référence : UMR8529-AURDUC-009
Nombre de Postes : 1
Lieu de travail : VILLENEUVE D ASCQ
Date de publication : samedi 29 novembre 2025
Type de contrat : IT en contrat de projet
Durée du contrat : 15 mois
Date d'embauche prévue : 1 mars 2026
Quotité de travail : Complet
Rémunération : between €2,491 and €2,695 gross monthly, depending on experience
Niveau d'études souhaité : BAC+5
Expérience souhaitée : Indifférent
BAP : F - Culture, Communication, Production et diffusion des savoirs
Emploi type : Ingenieure ou ingenieur des systemes et techniques audiovisuels et multimedia
Missions
- As part of the ANR project “IDANOPO – Identity and anonymity in portraiture from 1600 to 1800: between art history and artificial intelligence,” design and implement all or part of the protocols for collecting, structuring, and processing digital visual corpora intended for training computer vision models.
- Design and create the scientific animation of the project: website and research notebook, coordination of workshops and events.
Activités
- Participate in the creation of the project's supervised training corpus (“field truths”).
- Create and annotate the training corpus for the unsupervised portrait recognition model.
- Ensure the implementation of open science principles within the framework of the ANR project (data storage in appropriate repositories).
- Participate in the development of the project's data management plan.
- Monitor scientific and technological developments related to the project's themes.
- Participate in project communication (website creation, scientific blog posts).
Compétences
- Proficiency in an object-oriented programming language (Python, etc.)
- Knowledge of the IIIF format
- Knowledge of how APIs work
- Knowledge of ontologies and open data and metadata formats for iconographic corpora
- Knowledge of the principles of open science
- Knowledge of art history (17th and 18th centuries)
Résultats attendus et contrôles
NA
Contexte de travail
The IDANOPO project aims to compare artificial intelligence (AI) facial recognition with art portraits from the 17th and 18th centuries in order to identify the people depicted and place them in their historical context. It brings together researchers in art history and computer-based digital image analysis to design a new tool for the historical study of portraiture and the circulation of models. The aim is to use computer vision to automatically analyze large corpora, only part of which will be supervised, as expert annotations are particularly time-consuming.
This project, which closely combines the humanities and computer science, also incorporates a strong digital humanities component in order to be part of a fully open science approach. It is organized around three areas of work and one cross-cutting area: the construction of learning corpora (area 1); the design of GNNs for automated corpus analysis (area 2); the analysis and interpretation of methods and results (area 3); and the communication and sharing of data via open science (area 4). The engineer will work mainly on areas 1 and 4 of the project and will be involved in areas 2 and 3.
Coordinated by Gaëtane Maës, the project brings together three teams attached to the HARTIS and CRIStAL research units at the University of Lille and the National Library of France, respectively.
Contraintes et risques
NA