PCA Shenanigans and How to ML | Jakob Ringler | Houdini Horizon

Learn about Principal Component Analysis (PCA) and its unconventional use to de-jitter geometries and volumes. Dive into two machine learning projects entirely built inside of Houdini, which make use of PCA and the new ML nodes introduced in 20.5. Learn how an ML-based groom deformer works and how you can train Houdini to understand your sketches.

Jakob Ringler is currently an FX Intern at SideFX Software. Before joining the company he worked in the VFX industry in Berlin, Germany as an FX TD. He contributed to many high profile commercials and was involved in a broad range of projects and tasks covering FX work, On-Set LED wall visuals, compositing and pipeline development. He is fascinated by the intersection of film, design and tech and is currently exploring the more technical aspects of the industry and expanding his knowledge of machine learning.