Gianni Franchi

Gianni Franchi is an assistant professor at ENSTA Paris, affiliated with the Computer Science and Systems Engineering Unit (U2IS). His work focuses on theoretical deep learning, with a strong emphasis on uncertainty quantification, robustness, and explainability in machine learning models.

• Current affiliation: ENSTA Paris (U2IS)
• Academic rank: Assistant Professor
• Key collaborators: David Filliat, Emanuel Aldea, Andrei Bursuc, Antoine Manzanera

His research spans uncertainty quantification, explainable AI, and reliable machine learning. He investigates methods like Bayesian neural networks, ensemble approaches, and deterministic uncertainty models. His work also addresses domain adaptation, self-supervised learning, and autonomous systems, particularly in trajectory forecasting and semantic segmentation for autonomous driving.

Recent publications analyze probabilistic modeling for robustness, symmetry-aware Bayesian methods, and multi-modal datasets like InfraParis. He develops frameworks like Torch-Uncertainty and benchmarks such as MUAD for uncertainty types in autonomous driving.

• Key themes:
  • Uncertainty Quantification
  • Deep Learning Theory
  • Autonomous Systems
  • Explainable AI
  • Dataset Creation
  • Bayesian Methods