networking app
Lessons and implications of moving from FACS to Machine Learning for Digital Human Acting
SessionLessons and implications of moving from FACS to Machine Learning for Digital Human Acting
DescriptionThere are many aspects as we explore and develop digital human pipelines built on Machine Learning (ML) compared to more traditional CGI FACS based approaches. Many of these issues are technical and after reviewing these technical approaches we will move beyond the technical to explore the implications for acting and actors when deploying ML such as Neural Rendering, VAEs(Deepfakes) and GANs. Using technical examples the talk will discuss some of the obligations we have in our relationship with actors, issues of data ownership and allowing the creative team to maintain authorship and direction of the performances in this new 'open green fields' area of ML driven digital humans.
Presenter
Mike Seymour
Co-Director
Motus Lab
The University of Sydney
Event Type
Featured Sessions
TimeThursday, 8 December 202211:15am - 12:00pm KST
LocationAuditorium, Level 5, West Wing
Languages
Formats
Registration Categories
Next PresentationNext Presentation
Technology Behind Mesh to Metahuman