Prototypage d'un réseaux de neurones sur une cible embarquée

• Prototyping of a neural network
• Development of a neural network on an embedded target (Raspberry Pi or T2080 with TensorFlow API)
• Dimensioning a neural network
• Performing neural network training
• Applying formal verification methods on the neural network
• Applying optimization techniques such as pruning and quantization on the neural network
• Deploying the optimized solution (C language) on the target
• Proposing a test environment and scenario to evaluate performance gains in execution time and memory consumption
• Production of source code and executable
• Documentation

• Student in Engineering School or Master 2 with specialization in embedded systems
• Mastery of C language
• Knowledge related to neural networks
• Knowledge of optimization (complexity)
• Knowledge of matrix calculation
• Solid skills in mathematics would be a plus

Solid skills in mathematics

• Enriching employee experience with flexible work environment
• Complete integration program
• Academy for skills development
• Strong societal commitment with over 3,000 supported associations
• Active disability policy
• Inclusive culture
• Responsible digital approach
• Partnerships like Elles bougent to inspire female talents