Parsa Esmati

关于

I’m a PhD student at University of Bristol. I work on the intersection of computer vision and physical models. My research specifically focuses on spatiotemporal generative modelling, as well as representation learning, with the objective of improving temporal consistency, and the physical understanding in generative vision models. I also have a background in numerical simulations for PDEs.

At Microsoft, I work with the game intelligence group on controllable world models, building robust and consistent generative world models with longer memories.