Nicolas Papadakis

Nicolas Papadakis is a Director of Research in Applied Mathematics at the Bordeaux Institute of Mathematics (IMB), University of Bordeaux, where he also serves as deputy director since January 2023. His work bridges advanced mathematical theory and practical AI applications, particularly in image analysis and processing.

His research focuses on developing mathematically robust neural networks to enhance the reliability of image recognition systems, especially under signal perturbations such as noise, blur, or brightness changes. He specializes in denoising neural networks and their deployment in critical domains like medical imaging, remote sensing, and biological data analysis.

The most recent publications reflect a strong trend in integrating optimization theory with deep learning, particularly in plug-and-play architectures and minimal-supervision learning for medical diagnostics. His work combines theoretical rigor with real-world impact, especially in oncology, where accurate 3D tumor reconstruction from MRI scans enables improved monitoring and treatment evaluation.

• Expert in mathematical foundations of AI and neural networks
• Focus on robustness and reliability in image recognition
• Applications in medical imaging, particularly brain tumor analysis

Nicolas Papadakis actively contributes to public understanding of AI through media engagements and science communication. He has participated in the 'Expert’s Word' series and co-authored educational articles on image colorization. While current grants and advising roles are not listed, his leadership position at IMB suggests involvement in major research initiatives and collaborative projects.

He has published in high-impact venues including ICLR, Medical Image Computing and Computer Assisted Intervention (MICCAI), and SIAM Journal on Imaging Sciences. His work is characterized by a deep integration of applied mathematics, optimization, and machine learning.