As a "forensic science", forensic anthropology must meet strict methodological constraints with regard to the stakes of the conclusions drawn. The reliability of analytical processing chains of remains (skeletonized or not) which leads to the estimation of biological profile presented in a court of law is part of it, and an interdisciplinary approach integrating recent advances in the field of Artificial Intelligence and machine learning is able to bring significant advances.
This project is built in collaboration with the LAENNEC Institute (Digital Sciences and Artificial Intelligence for Health) and the LIS UMR 7020 (Computer Science and Systems Laboratory). The stakes for our discipline are absolutely strategic. Just as medicine, which has long been an "art", is now being turned upside down by Data Science, data analysis and AI are destined to transform the world of forensic science based on the analysis of biological data. The objective of this project is therefore to improve biological profile estimates via robust statistical methods and machine learning algorithms (which have already proven their worth in the high-tech industry), and to develop a "global validity indicator" of the estimate presented during the expert's deposition in court. An inter-doctoral school thesis has just begun on this topic (October 2023) and an ERC project is being prepared (submission expected in 2024). The " Forensic Anthropology Society of Europe - FASE " has recently published a statement to promote research that elevates forensic anthropology to the rank of "primary" identifier in the eyes of international courts of justice, and as such has agreed to be a partner in this project and in the envisaged ERC submission.