3D Mesh Skinning With Global-Local Joint Weight Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing skinning methods suffer from low accuracy due to manual binding of vertices to joints, resulting in errors in the binding relationship between vertices and joints.
Innovation Solution
A method involving global and local joint weight prediction using neural network models to optimize joint weight information, combining global and local predictions to improve accuracy, particularly for complex costumes in 3D objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual binding of vertices to joints is used, then the process is simple and controllable, but the accuracy of joint weight prediction is low
Solution Approach 1:
The skinning process is divided into two stages: global joint weight prediction for all vertices, and local joint weight prediction for target vertices that need optimization. This segmentation allows the system to achieve high accuracy through local refinement while maintaining overall efficiency through global prediction, resolving the contradiction between accuracy and complexity.
Solution Approach 2:
The system performs preliminary global joint weight prediction for all vertices before conducting local prediction for specific target vertices. This preliminary action establishes a baseline that improves the efficiency and accuracy of subsequent local optimization, addressing the contradiction by preparing comprehensive data upfront.
2Productivity
If global joint weight prediction is performed on the entire 3D mesh, then the processing efficiency is improved, but the accuracy for complex costumes deteriorates
Solution Approach 1:
The prediction process is segmented into global prediction for all vertices (providing efficiency) and local prediction for target vertices (providing accuracy). This allows the system to maintain high processing efficiency through global operations while achieving high accuracy through targeted local refinement, particularly for complex costumes.
Solution Approach 2:
The system applies different prediction qualities to different parts of the mesh: global prediction is applied uniformly to all vertices for efficiency, while local prediction with optimization is applied only to target vertices where higher accuracy is needed, such as complex costume regions. This local quality approach resolves the contradiction between efficiency and precision.
3Measurement precision
If local prediction is performed on target vertices, then the accuracy of joint weight information is improved, but the computational time increases
Solution Approach 1:
Instead of performing local prediction on all vertices (excessive action), the system performs local prediction only on target vertices that require optimization (partial action). This partial approach achieves the necessary accuracy improvement while minimizing computational time by avoiding unnecessary processing on vertices where global prediction was already sufficient.
Solution Approach 2:
The system performs preliminary global prediction on all vertices before identifying target vertices for local prediction. This preliminary action establishes accurate baseline joint weight information that reduces the computational burden and time required for subsequent local optimization, resolving the contradiction between accuracy improvement and time loss.
Data Source
AI summary
A skinning method includes: acquiring a 3D mesh for skinning of a skeletal structure of a 3D object, the 3D mesh including a plurality of vertices and the skeletal structure including a plurality of joints; respectively acquiring global joint weight information corresponding to the plurality of vertices, the global joint weight information of a vertex including associated joints of the vertices predicted based on the 3D mesh, and joint weight values of the associated joints; determining a target vertex to be locally predicted from the vertices of the 3D mesh based on the global joint weight information; performing a local prediction based on a vertex feature of the target vertex to obtain local joint weight information of the target vertex; and optimizing global joint weight information of the target vertex based on the local joint weight information for skinning of the 3D object using the optimized joint weight information.