3D Generative Models (23w5153)


Michael Bronstein (Oxford University)

(Simon Fraser University)

(The University of Pennsylvania)

(University of Washington)

(SFU, Google Deepmind)

(University of Guelph)


The Banff International Research Station will host the "3D Generative Models" workshop in Banff from July 9 to July 14, 2023.

Learning from large collections of images, generative models implemented by neural networks are currently able to synthesize photo-realistic images with some degree of control over different aspects of style.

Simultaneously, enormous progress has also been made in the representation of 3D information with neural networks: current technology allows us to encode multiple images of a 3D scene within the parameters of an artificial neural network and then artificially observe the 3D scene from a new point of view.

These two research thread have evolved mostly independently within their own sub-community, hence in this workshop we will attempt to bridge the gap in these developments by asking two core questions:

1. How can we leverage generative models to handle the inherent uncertainty in the 3D structure estimated from only 2D images?

2. How should generative models be improved to meet the requirements induced by the estimation of 3D structure?

In this workshop, we will explore how the research realms of generative modelling and 3D neural representations can benefit from each other, as well as explore novel applications in different areas of science and technology.

The Banff International Research Station for Mathematical Innovation and Discovery (BIRS) is a collaborative Canada-US-Mexico venture that provides an environment for creative interaction as well as the exchange of ideas, knowledge, and methods within the Mathematical Sciences, with related disciplines and with industry. The research station is located at The Banff Centre in Alberta and is supported by Canada's Natural Science and Engineering Research Council (NSERC), the U.S. National Science Foundation (NSF), Alberta's Advanced Education and Technology, and Mexico's Consejo Nacional de Ciencia y Tecnología (CONACYT).