Descriptif

Generative AI enables machines to interpret and process data like humans, via multiple modalities and sensors like vision, audio, tactile information. In the constantly evolving field of AI, the emergence and rapid development of Generative AI with multiple modalities is a major scientific advancement. However, this development of multimodal Generative AI is not only a scientific advancement; it is a paradigm shift that is reshaping industries, from healthcare to automotive, and changing how we live every day. For instance, new tools like Dall-E 3, Midjourney, and GPT-Vision are all leveraging large vision models, while autonomous driving companies (such as Uber, Tesla, etc) are expanding. 

This class will go beyond CNNs and transformers and cover the advanced topics in multimodal generative AI: from image to video generation, sound, robot actions.