Pose-Space Shape Fitting via Clustering and Bid Points
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current techniques for interpolating the shape of objects in computer animation are inefficient, especially for complex objects with many animation variables, leading to computationally prohibitive processes and undesirable shapes.
Innovation Solution
The method involves clustering training poses, determining a bid point for each cluster, and using these points to calculate a cluster-fitted shape for the desired pose through a linear combination of basis vectors, allowing for efficient shape fitting even with limited training poses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If current interpolation techniques are used for complex objects with many animation variables, then shape interpolation can be performed, but the computational complexity becomes prohibitive
Solution Approach 1:
The patent segments the high-dimensional pose space into multiple lower-dimensional subspaces by identifying independent animation variable groups. Each subspace is processed separately with its own interpolation, reducing the overall computational complexity while maintaining shape interpolation accuracy across the full pose space.
Solution Approach 2:
The patent transforms the high-dimensional pose space interpolation problem into multiple lower-dimensional subspace problems. By changing the dimensionality approach and processing each subspace independently, the computational burden is reduced while preserving the ability to interpolate shapes accurately in the original high-dimensional space.
2Measurement precision
If the number of training poses is increased to cover more of the pose space, then interpolation precision improves, but the computational cost increases significantly
Solution Approach 1:
The patent segments the pose space coverage requirement into multiple subspaces, each covered by a smaller set of training poses. This allows adequate coverage of the entire pose space using fewer total training poses than would be required to uniformly cover the full high-dimensional space, improving computational efficiency while maintaining interpolation precision.
3Shape
If manual shape adjustment is performed for each training pose to achieve desirable shapes, then shape quality improves, but the time and effort required increases
Solution Approach 1:
The patent performs preliminary clustering of training poses and identification of independent animation variable groups before the shape adjustment process. This preliminary organization reduces the number of poses that require manual adjustment and structures the data to minimize iterative refinement, thereby improving shape quality while reducing the time and effort required.
Data Source
AI summary
Techniques relate to fitting a shape of an object when placed in a desired pose. For example, a plurality of training poses can be received, wherein each training pose is associated with a training shape. The training poses can be clustered in pose space, and a bid point can be determined for each cluster. A cluster-fitted shape can then be determined for a pose at the bid point using the training shapes in the cluster. A weight for each cluster-fitted shape can then be determined. The cluster-fitted shapes can then be combined using the determined weights to determine a shape of the object in the desired pose.