Multidisciplinary Institute in Artificial intelligence

Synopsis

In this presentation, I will discuss image generation, also known as visual generative AI. I will provide an overview of the state-of-the-art approaches for generating images and videos. In particular, I will present two classes of approaches in detail, highlight open problems associated with these methods, and finally discuss ongoing research directions to address them.

Speaker

Karteek Alahari is a Research Director at the Inria Center of Université Grenoble Alpes. Before joining Inria in 2010, he conducted research in the United Kingdom, India, and the United States. His main research interests focus on robust and efficient learning of visual representations (from images and videos), particularly when data is partially labeled. This includes frameworks such as incremental learning, weakly supervised learning, multimodal learning, and more.

Replay the webinar