Virtual Avatar Mesh Matching with Regional Rigid and Non-Rigid Deformation
Solution Overview
Problem
Existing systems face challenges in efficiently creating high-fidelity virtual avatars for augmented, virtual, and mixed reality environments, requiring extensive resource-intensive scanning processes and lacking accurate animation techniques.
Innovation Solution
A method for automatically matching a base mesh of an animation rig to a target mesh using rigid transformations in regions of higher error and non-rigid deformations in regions of lower error, iteratively refining the match until an error tolerance is met, to generate blendshapes for various poses of avatars.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If extensive scanning processes are used to create high-fidelity virtual avatars, then avatar accuracy and realism are improved, but resource consumption and processing time increase significantly
Solution Approach 1:
The patent segments the mesh matching process into multiple iterative cycles, each focusing on different regions of the avatar model. Base meshes are divided into multiple target meshes representing different poses, and the matching process is broken down into successive refinement steps that progressively improve accuracy while maintaining computational efficiency.
Solution Approach 2:
The patent performs preliminary actions by pre-processing the base mesh into multiple target meshes in advance, and by using rigid transformations as an initial matching step before applying more computationally intensive non-rigid deformations. This preliminary structuring reduces the overall computational burden during the main matching process.
2Manufacturing precision
If traditional mesh matching methods are used, then comprehensive coverage of the avatar model is achieved, but processing time and computational resources increase
Solution Approach 1:
The patent divides the avatar mesh into multiple regions and processes them through successive iterative cycles. Each cycle focuses on matching specific target meshes to the base mesh, allowing the system to make targeted improvements without reprocessing the entire model, thus reducing overall processing time while maintaining comprehensive coverage.
Solution Approach 2:
The patent implements periodic action through iterative refinement cycles that repeatedly apply transformation algorithms to progressively improve mesh matching accuracy. Each iteration builds upon the previous one, allowing the system to converge to high accuracy solutions through multiple passes rather than requiring a single computationally expensive operation.
3Manufacturing precision
If rigid transformations are applied to high-error regions, then matching accuracy in critical areas is improved, but overall model flexibility decreases
Solution Approach 1:
The patent applies local quality by using rigid transformations specifically for high-error regions where structural accuracy is critical, while employing non-rigid deformations in lower-error regions that require more flexibility. This spatially varying transformation strategy optimizes both local precision and global adaptability of the mesh matching process.
Solution Approach 2:
The patent implements dynamics by allowing the transformation type to vary dynamically across different regions and iterations. The system adaptively selects between rigid and non-rigid transformations based on error levels and region characteristics, enabling the matching process to be both precise where needed and flexible where appropriate.
Data Source
AI summary
Examples of systems and methods for matching a base mesh to a target mesh for a virtual avatar or object are disclosed. The systems and methods may be configured to automatically match a base mesh of an animation rig to a target mesh, which may represent a particular pose of the virtual avatar or object. Base meshes may be obtained by manipulating an avatar or object into a particular pose, while target meshes may be obtain by scanning, photographing, or otherwise obtaining information about a person or object in the particular pose. The systems and methods may automatically match a base mesh to a target mesh using rigid transformations in regions of higher error and non-rigid deformations in regions of lower error.