Lucio Moser is the R&D Software Lead in the Digital Humans Group at Digital Domain where he works on facial motion capture and mapping, machine learning, geometric processing and computer vision.
Here, Lucio provides an excellent insight on what goes on inside the Digital Humans Group’s R&D environment, where they have been using machine learning to solve the ‘uncanny valley’ problem of creating photorealistic digital humans. This video focuses on the redesigned facial motion capture workflow that produced Masquerade 2.0. and how machine learning is making us rethink the typical approaches to old problems and letting us solve problems that seemed insurmountable.
Source: The VFXRio Vimeo page