Stage - Optimisation de l’exécution de modèles d’IA pour les systèmes robotiques - Industrie - Toulouse
Core
Design and evaluate a scheduling strategy to optimally execute multiple concurrent DNN models for advanced perception and tracking functions in robotic systems, considering activation periods, deadlines, priorities, and computational resources.
Role type
Intern, Embedded Systems & AI Scheduling
Builds
Real-time robotic perception and tracking systems with dynamic model activation/deactivation
Domain
Robotics, Embedded Systems, Real-Time Systems, Artificial Intelligence
Deliverable
production ML models
Required skills
DNN modeling, Real-time scheduling algorithms, Python, C++, Resource allocation (CPU/NPU), Latency optimization
Preferred skills
Multi-DNN inference chains, Dynamic policy adaptation, Heterogeneous computing architectures
Technologies
DNN, Python, C++, CPU, NPU
Responsibilities
Model inference chains and temporal constraints, Dynamic scheduling strategies, Operational context adaptation, Resource allocation optimization, Demonstrator implementation and validation
Seniority
Intern
