Post-doctoral researcher in bias evaluation and mitigation for 3D facial reconstruction

Organisation/Company IMT - Institut Mines-Télécom Research Field Computer science » 3 D modelling Researcher Profile First Stage Researcher (R1) Positions Postdoc Positions Application Deadline 20 Nov 2027 - 23:59 (Europe/Paris) Country France Type of Contract Temporary Job Status Full-time Offer Starting Date 1 Mar 2027 Is the job funded through the EU Research Framework Programme? Not funded by a EU programme Is the Job related to staff position within a Research Infrastructure? No

Offer Description

A leading general engineering school within IMT (Institut Mines-Télécom), the number one group of engineering schools in France, IMT Atlantique aims to support transitions and train responsible engineers. We place scientific and technical excellence at the service of education, research, and innovation.

As part of the ANR-DFG 3DFaceBias project, which aims to measure, explain and reduce accuracy disparities when reconstructing a 3D face from 2D photographs, we are hiring a Post-Doctoral Researcher in bias evaluation and mitigation for 3D facial reconstruction within Data Science department of IMT Atlantique and in close collaboration with Bernhard Egger, the German principal investigator, and his team at FAU.

For the French–German 3DFaceBias project, and under the scientific supervision of Sarah Matta, the French principal investigator, and Tinashe Mutsvangwa, you will lead the IMT Atlantique research on bias evaluation and mitigation for 3D facial reconstruction.

Responsibilities

• Audit bias in 3D face reconstruction from 2D images.
• Identify the sources of bias through data-centric analysis.
• Design and evaluate data-centric strategies to mitigate bias.
• Contribute to cross-cutting project activities.

Qualifications

• Strong experience in empirical machine-learning research and advanced statistical methods.
• Rigorous experimental design: controlled comparisons, uncertainty, confounding, distribution shifts and prevention of data leakage.
• Proficiency in Python and a machine-learning framework such as PyTorch; ability to modify, test and debug research code.
• Computer vision, 3D reconstruction, 3D geometry, statistical shape models or medical image computing.
• Algorithmic fairness, robustness, generalisation, dataset preparation or synthetic image generation.
• Performance and bias evaluation: subgroup analyses, interpretation of disparities and careful assessment of limitations.
• Independence, scientific initiative and the ability to formulate hypotheses and prioritise experiments.
• Rigour, reproducibility, documentation and collaborative management of code and data.
• Effective international teamwork and clear communication of findings, limitations and progress.

Previous expertise in 3D facial reconstruction is an asset. Applicants with strong empirical machine-learning and statistical backgrounds are encouraged to apply; specialist geometry and reconstruction skills can be developed during the project.

Specific Requirements

Good written and spoken scientific English, including strong academic writing skills.

Candidates must hold a Ph.D. obtained less than 3 years prior to the hiring date

Languages ENGLISH Level Good

Additional Information

Take scientific ownership within an international project on the reliability and fairness of 3D reconstruction methods.

Work at the intersection of machine learning, statistics and medical image computing.

Develop your experimental research and publications within a French–German collaboration with complementary expertise.

Advance responsible and reproducible research through transparent evaluation, careful handling of facial data, and reusable methods and code shared wherever permissions allow.

BENEFITS

• Leisure and sports offers
• Public transport subsidies
• Wide range of social benefits
• Generous leave package

Eligibility criteria

Candidates must hold a Ph.D. obtained less than 3 years prior to the hiring date

#J-18808-Ljbffr