Neural Video Style Transfer Leveraging PDG | Manuel Koster | Houdini Hive Worldwide

Neural Style Transfer is a method of transferring a style of one image to another image using Machine Learning. Video Style Transfer also includes temporal consistency – meaning, there is no flickering inbetween frames.This project focusses on leveraging PDG to distribute work needed for the stylisation. It also gives an example of how PDG can be used to parameterize all kinds of processes to increase availability to less technical artists out there.

Manuel Köster is a Technical Artist/Director who has been working in VFX/Animation with movies such as Dr. Strange, Cpt. Marvel, Lego Movie 2 as well as Peter Rabbit 2. He recently switched to the Game industry and joined Crytek in an effort to enhance their procedural workflows. He has also been teaching Houdini for his previous University and online with his website shadesoforange.de

Leave a Reply

Your email address will not be published. Required fields are marked *