BA HONORS PROJECT
This dissertation project focused on the research and development of an advanced, realistic facial rig for use in both hand-keyed animation and motion-capture data integration (real-time and non-real-time). The research encompassed a comprehensive analysis of current markerless facial motion capture technologies, including both real-time and non-real-time solutions, and an in-depth evaluation of advanced rigging techniques, such as 3 Lateral, MetaHuman, and Faceware systems.
The primary objective of the project was to design and implement a highly realistic facial rig that could seamlessly accommodate both hand-keyed and motion-captured animation workflows.
RIG & ANIMATION
The facial rig was created and animated in Autodesk Maya using a combination of bone-based rigging, rivet commands, and blendshapes, following the methodology of the Facial Action Coding System (FACS) for precise facial muscle control.
A custom Faceboard was developed utilizing the Node Editor and Set Driven Keys, allowing for streamlined control and efficient animation workflows.
MOCAP WITH FCLONE
The facial rig is seamlessly integrated with FClone software using a combination of blendshapes and shapers, following the Facial Action Coding System (FACS). For motion capture, I employed the Microsoft Kinect 2 to record facial performance, which was subsequently processed with overlay markers that I set up specifically for my facial structure within FClone.
This integration allows for the real-time streaming or export of captured data, enabling direct import into the chosen Digital Content Creation (DCC) package for further animation and refinement.