Texture Joint Animation With Offline Binding for Multi-Character Scaling

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing animation techniques, such as skeletal mesh animation, struggle to scale well for animating multiple characters simultaneously due to high central processing unit utilization, while texture vertex animation lacks runtime flexibility and is limited to specific characters and levels of detail.

Innovation Solution

Texture joint animation involves precomputing skeletal animation data and mesh binding data offline, allowing for runtime deformation on a graphics processing unit, enabling reuse of the same skeleton and animation with different characters and levels of detail, and providing runtime selection of animation subsets.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If skeletal mesh animation is used to animate multiple characters simultaneously, then runtime flexibility is maintained, but central processing unit utilization becomes excessively high and scaling is poor

Engineering Contradiction:
Improveruntime flexibilityVSAvoidscaling to multiple characters
Core Design Contradiction:
Adaptability or versatilityVSProductivity

Solution Approach 1:

The patent segments the animation processing into distinct components: skeletal animation data generation (offline), mesh binding data creation (offline), and runtime deformation execution (GPU). This segmentation allows CPU-intensive preprocessing to be separated from runtime rendering, enabling better scaling to multiple characters while preserving runtime flexibility through GPU-based deformation.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary actions by precomputing skeletal animation data and mesh binding data offline before runtime. This preliminary processing reduces the computational burden during runtime, allowing the system to scale to multiple characters efficiently while maintaining the ability to apply arbitrary animations through runtime deformation control.

Inventive Principle:
Principle #10Preliminary action

2Productivity

If texture vertex animation is used for concurrent animation of multiple characters, then performance is improved, but the technique is limited to specific characters at specific levels of detail and lacks runtime variation

Engineering Contradiction:
Improveconcurrent animation performanceVSAvoidcharacter and level of detail flexibility
Core Design Contradiction:
ProductivityVSAdaptability or versatility

Solution Approach 1:

The patent creates a universal animation system where skeletal animation data and mesh binding data can be applied to multiple characters at different levels of detail. The mesh binding data establishes flexible mappings that allow the same skeletal animation data to drive different meshes, enabling one animation set to serve multiple characters and LODs without recompute.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent enables runtime variation by allowing dynamic modification of deformation parameters during execution. The system computes vertex positions based on skeletal transformation data that can be adjusted at runtime, permitting animation variation without recompilation and supporting different characters and LODs through parameter changes rather than fixed precomputation.

Inventive Principle:
Principle #35Parameter changes

3Productivity

If texture vertex animation precomputes the entire animation, then performance is improved, but runtime variation of the animation is not allowed

Engineering Contradiction:
Improveanimation computation efficiencyVSAvoidruntime animation variation
Core Design Contradiction:
ProductivityVSAdaptability or versatility

Solution Approach 1:

The patent performs preliminary computation of skeletal animation data and mesh binding data offline to improve runtime performance, but结构设计 allows runtime variation by separating the precomputed binding relationships from the actual deformation application. This enables efficient runtime modification of animation parameters while maintaining the performance benefits of precomputation.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent introduces dynamics by enabling runtime modification of animation parameters and deformation control. The system transitions from static precomputed animations to dynamic runtime-controllable deformations, allowing animation variation while maintaining computational efficiency through the preestablished mesh binding framework.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP4648014A1Texture joint animation
Publication Date: 2025.11.12 MICROSOFT TECHNOLOGY LICENSING LLC
  • EP4648014A1 patent drawingFigure 1
  • EP4648014A1 patent drawingFigure 2A~2C
  • EP4648014A1 patent drawingFigure 3

AI summary

The description relates to computer graphics techniques for generating animations. In the disclosed techniques, skeletal animation data is generated representing a skeleton. A mesh is then bound to the skeleton to create mesh binding data. The skeletal animation data and the mesh binding data can be used to deform the mesh to generate an animation. In some cases, the skeletal animation data and the mesh binding data can be generated during an offline process. Then, the deformation can be performed at runtime, e.g., by a graphics processing unit. The animation can be implemented in many different types of applications, including video games, virtual reality applications, augmented reality applications, etc.