Automated Human Synthesis: Photoreal Text to Digital Human | Conor Collins & Matt Benson | Houdi...

Synthesis AI brings together generative AI and traditional Houdini procedural workflows to create a flexible AWS cloud-based pipeline capable of infinite automated asset and synthetic data production. Today we share an overview of the pipeline architecture and our internal tools used create photoreal digital humans.
Conor Collins is a versatile Technical Artist with over 15 years of experience across multiple careers and disciplines in digital imaging. Beginning his journey in photography and now immersed in generative AI and Houdini, Conor's passion for digital artistry propelled him into new realms of CG, motion design, and generative work. His ability to blend technical innovation with creative problem-solving has been a hallmark of his career, leading to his current role at Synthesis AI, where he continues to push the boundaries of digital imaging through his diverse expertise.
Matt Benson is a seasoned technical artist with over 10 years of experience in procedural workflows across various domains. Beginning his career at Pixar Animation Studios, Matt contributed to 11 feature films as well as several short films across multiple departments. He brings his passion for procedural problem-solving to Synthesis AI, where he works at the forefront of artificial intelligence and computer vision.

[@mattestela]

amazing talk, thanks for sharing all that info!

[@uzivatelInternetu]

Hello, as senior FX artist I would like to ask, what are possibilitites of use Machine Learning in simulations. I can see so many AI generative models around, which are already capable to create amazing fires, smokes, explosions etc. So I can imagine that in following years we FX artists could stay without jobs. :) Thank you for insight.