Computer vision and deep learning - Part 2

Content

The goal is to create an artificial intelligence solution dedicated to a specific application. The different types of networks will be discovered in a practical way through the handling of existing solutions corresponding to a set of applications (examples: object recognition, face recognition, 2D/3D reconstruction, text detection and recognition, scene segmentation, identification of objects of interest speech recognition...). In this context, the student will be required to install and configure the tools for deploying neural networks (Tensorflow, Pytorch...), then experiment and evaluate the approaches considered. A work in pairs will be set up. In addition to the issues proposed by the teaching team at the beginning of the module, students are encouraged to propose topics that interest them. Plenary sessions of restitution will be periodically set up in order to mutually enrich the pedagogical experience.