Missions

Scientific background
Anthracology is a robust method for studying forest stands and their transformation as a result of climate change or human practices, as well as the use of wood as a fuel and material. This method is based on the botanical identification of large bodies of wood and charcoal preserved on archaeological sites. It is based on a visual or morphometric microscopic reading of the anatomical structure of the wood, preserved through carbonization. The quality of the work relies on the researchers' ability to identify taxa. While the specialist's expertise remains effective for the identification of most taxa, the anatomical proximity of certain species remains a barrier to the identification of other taxa with high information value. By drawing on the latest advances and developments in AI research, and on new developments, the aim of the AI-WOOD project is to propose, via learning models, a decision-making tool for the identification of taxa for which conventional methods are inoperative, even though the archaeological and paleo-environmental issues associated with their identification are numerous (mainly Gymnosperms, Ericaceae and Rosaceae). By enabling the identification of key taxa, the project will provide unprecedented information on the evolution of forest stands and their uses, from the Paleolithic to the sub-current. At the end of the project, a royalty-free bank of labelled images and an interoperable interface will be made available to the wood science community at large. It may also find applications in industry for species recognition.

Job description:
The candidate must be an anthracologist, and be able to grasp the principles of machine learning, without necessarily mastering the technical aspects and/or those linked to fundamental research. He/she will participate in two phases of the project:
Phase 1) learning data: He/she will work closely with the Inria AI team, and more particularly with the PhD student in charge of developing the algorithm from the learning database (images of actual wood). Together, they will define the specificities linked to wood anatomy (in particular taxonomy, anatomical areas of interest, as well as the nature and quality of the images).
Phase 2) The post-doctoral researcher will actively participate in the test phase, using archaeological charcoal from European sites, covering a chronology from the Paleolithic to the Middle Ages. The focus will be on taxa such as gymnosperms, ericaceae and rosaceae, collected (as images or samples to be photographed) from project members. He/she will evaluate the relevance of the taxonomic attributions proposed by the algorithm and, on this basis, formulate proposals for identification and archaeobotanical interpretation.

Activités

- Collect existing wood anatomy photographs from the anthracology community
- Take anatomical photographs (by light and electron microscopy) of archaeological wood charcoals from Gymnosperms, Ericaceae and Rosaceae.
- Collaborate in the implementation of algorithms adapted to the anatomical characteristics of the wood
- Test the validity of algorithm outputs
- Participate in team meetings and field activities
- Propose identifications for charcoal not identified by conventional methods.

Compétences

Knowledge :
- Precise knowledge of wood anatomy
- Knowledge of ecology
- General knowledge of machine learning principles
- General knowledge of safety and hygiene rules applicable to activities

Know-how :
- Use of optical microscopes (reflection and transmission)
- Use of scanning electron microscope (SEM)
- Anatomical photography

Personal skills :
- Organize day-to-day work
- Working in a team - Reporting on activities

Contexte de travail

Reporting to Isabelle Théry-Parisot, project leader, the post-doctoral fellow will work within the framework of ANR AI-WOOD. He/she will be based at CEPAM in Nice, within the IAMAHA team, and will actively participate in meetings and field activities.

Starting date: between 07/15/2025 and 09/15/2025

Contraintes et risques

Safety and hygiene rules applicable to anthropological observation in the laboratory and sampling/identification in the field