Head Pose Neutralization for Automatic Blend Shape Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for producing corrected neutral poses for character meshes are time-consuming and expensive, particularly when using skeletal animation based on human or entity captures.
Innovation Solution
A system that automatically corrects head poses and generates blend shapes by using rigging information, reference meshes, and transformation techniques to generate a neutral mesh, applying as-rigid-as-possible deformations and blend shape generation based on linear regression and position offsets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If manual correction of neutral poses is performed for meshes generated from captures, then pose accuracy is improved, but time consumption and cost increase
Solution Approach 1:
The system performs self-correction by automatically detecting pose deviations from neutral positions and computing correction transformations without requiring manual intervention. The correction module uses the mesh structure and capture data to autonomously generate corrected neutral poses, eliminating the need for time-consuming manual adjustment while maintaining pose accuracy.
Solution Approach 2:
The patent replaces manual mechanical adjustment with an automated computational system. Instead of physically manipulating mesh vertices to achieve neutral poses, the system uses algorithms to calculate transformation matrices that automatically correct pose deviations, substituting human-operated mechanical processes with automated computational methods.
2Manufacturing precision
If manual correction of neutral poses is performed for meshes generated from captures, then pose accuracy is improved, but production cost increases
Solution Approach 1:
The system performs self-correction by automatically detecting pose deviations from neutral positions and computing correction transformations without requiring manual intervention. The correction module uses the mesh structure and capture data to autonomously generate corrected neutral poses, eliminating the need for time-consuming manual adjustment while maintaining pose accuracy.
Solution Approach 2:
The patent replaces manual mechanical adjustment with an automated computational system. Instead of physically manipulating mesh vertices to achieve neutral poses, the system uses algorithms to calculate transformation matrices that automatically correct pose deviations, substituting human-operated mechanical processes with automated computational methods.
3Productivity
If automated head pose correction is implemented, then productivity is improved, but system complexity increases
Solution Approach 1:
The system divides the complex task of pose correction into distinct modular components: a detection module that identifies pose deviations, a correction module that computes transformation matrices, and a blend shape generation module that applies corrections. This segmentation allows each component to handle specific sub-tasks independently, improving processing efficiency while managing system complexity through modular architecture.
Solution Approach 2:
The patent introduces transformation matrices as an intermediary computational element that bridges the detection and correction stages. These matrices serve as a mathematical mediator that translates detected pose deviations into corrected mesh configurations, enabling automated correction without requiring direct complex manipulation of mesh geometry.
Data Source
AI summary
A system may perform head pose neutralization on an input mesh to produce a neutral mesh and/or determine blend shapes for the neutral mesh. The system may generate a neutral mesh based on an input mesh and a reference mesh and then generate a blend shape associated with the neutral mesh based at least in part on one or more reference neutral meshes and one or more corresponding reference blend shapes.