Virtual Avatar Facial Animation with Anatomical Subregion Solvers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems for creating high fidelity digital avatars in virtual reality, augmented reality, and mixed reality environments are resource-intensive and time-consuming, often requiring extensive scanning processes that are impractical for users, and lack efficient methods for animating facial expressions with biologically motivated results.
Innovation Solution
The use of a Facial Action Coding System (FACS) taxonomy integrated into a facial rig, where facial subregions are solved iteratively, starting from large areas like the jaw and progressing to granular details like the eyes, to create a biologically motivated and computationally robust avatar animation system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If extensive scanning processes are used to create high fidelity digital avatars, then the fidelity and accuracy of the avatar is improved, but the time consumption and resource requirements increase significantly
Solution Approach 1:
The patent segments the facial animation process into distinct subregion solvers that operate independently on different facial areas (e.g., jaw, eyes, mouth). Each subregion solver handles specific facial parameters, allowing parallel processing and reducing overall computation time while maintaining high fidelity through localized optimization.
Solution Approach 2:
The system performs preliminary actions by pre-defining facial subregions and their associated parameters before the actual scanning and animation process. This pre-organization of facial structure allows for more efficient processing during avatar creation, reducing the time required for high fidelity reconstruction.
2Manufacturing precision
If extensive scanning processes are used to create high fidelity digital avatars, then the fidelity and accuracy of the avatar is improved, but the resource consumption increases significantly
Solution Approach 1:
By dividing the facial animation system into multiple independent subregion solvers, the patent enables distributed and parallel computation. This segmentation reduces the computational burden on single processors, optimizes resource utilization, and allows for more efficient energy consumption during high fidelity avatar creation.
Solution Approach 2:
The system applies partial action by focusing computational resources on specific facial subregions that require higher fidelity rather than uniformly processing the entire face. This selective approach reduces overall resource consumption while maintaining the necessary level of detail in critical areas.
3Device complexity
If facial expressions are animated using traditional methods, then the computational simplicity is maintained, but the biological accuracy and realism of the animation deteriorates
Solution Approach 1:
The patent transforms the animation approach by changing parameters from global facial transformations to localized subregion-specific parameters. Each subregion solver adjusts parameters independently based on biological facial anatomy, achieving higher biological accuracy while maintaining computational tractability through parameterized models.
Solution Approach 2:
The system applies local quality by assigning different animation properties and parameter types to different facial subregions. Each region (eyes, jaw, mouth) has specialized solvers that apply locally-appropriate transformation rules, improving biological accuracy without requiring uniformly complex computation across the entire face.
4Productivity
If all facial subregions are solved simultaneously, then the computational directness is maintained, but the solution accuracy and convergence deteriorates
Solution Approach 1:
The patent divides the facial animation problem into segmented subregion solvers that process different facial areas independently. This segmentation allows each solver to converge more accurately on its specific parameters without the computational interference that would occur in a simultaneous all-at-once approach, while still maintaining overall system efficiency.
Data Source
AI summary
Systems and methods generating an animation ng corresponding to a pose of a subject include accessing image data corresponding to the pose of the subject. The image data can include the face of the subject. The systems and methods process the image data by successively analyzing subregions of the image according to a solver order. The solver order can be biologically or anatomically ordered to proceed from subregions that cause larger scale movements to subregions that cause smaller scale movements. In each subregion, the systems and methods can perform an optimization technique to fit parameters of the animation rig to the input image data. After all subregions have been processed, the animation rig can be used to animate an avatar to appear to be performing the pose of the subject.