Large generative vision models for content creation and representation learning

Saman Motamed (Ph.D. Student)

l'm broadly interested in Generative Vision models for content creation and representation learning. My research falls in to the following two categories: 1) I aim to gain a better understanding of how to enable user-intuitive control over generative models for personalized content creation with limited data. This includes topics such as using generative models for image and video editing, inpainting, and customizing the generated content based on user-specified criteria. 2) With generative models having advanced in realistically representing actions and object interactions, I am interested in exploring the representation learning ability of large generative vision models for downstream tasks where discriminative models have traditionally excelled (classification, depth estimation, etc). By working on both of these areas, I hope to improve how we create personalized content and expand the use of generative models in practical tasks.

Primary Host:	Luc Van Gool (ETH Zürich & KU Leuven)
Exchange Host:	Andrea Vedaldi (University of Oxford)
PhD Duration:	01 April 2023 - 01 April 2027
Exchange Duration:	01 April 2025 - 30 September 2025 - Ongoing