STAGE 2027 - Stage en Intelligence Artificielle pour le centre d'Essais en vol et intégration (F/H)
Core
Design, implement, and evaluate certification and uncertainty estimation approaches for critical AI models in computer vision used for flight test and integration measurement chains.
Role type
Intern, applied AI researcher (uncertainty quantification)
Builds
Certified AI models for flight test measurement systems
Domain
Aerospace, computer vision, algorithmic certification
Deliverable
production ML models
Required skills
Conformal prediction, distribution-free uncertainty quantification, Lipschitz neural networks, computer vision (OCR, image comparison), Python, PyTorch/TensorFlow, OpenCV, applied mathematics (probability, statistics, optimization)
Preferred skills
Experience in algorithmic robustness, industrial R&D projects, familiarity with aerospace/space/auto industries
Technologies
PyTorch, TensorFlow, OpenCV
Responsibilities
Conduct literature review on conformal prediction for deep neural networks, implement distribution-free uncertainty quantification, develop certified robustness techniques, benchmark approaches on industrial computer vision use cases, collaborate with test and physics teams to integrate solutions
Seniority
Intern, Master 2 or Engineering School 5th year