Avatar Carriage Format for Dynamic Volumetric Mesh Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for streaming animation data, such as those outlined in ISO/IEC 23090-14:2023, face limitations in accurately mapping and animating volumetric meshes due to unreliable correspondences between shadow and volumetric meshes, particularly in avatars with distinct geometry representations, and lack support for dynamic avatar animation and partial registration of mesh components.
Innovation Solution
A new carriage format using glTF 2.0 extensions, including MPEG_accessor_timed and MPEG_buffer_circular, enables timed-based updates of correspondences and pose information, allowing for dynamic mesh sequences and accurate mapping between different avatar representations, supporting features like avatar accessories, isolated body part animation, and haptic feedback.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If pre-computed correspondences between shadow mesh and volumetric mesh are used for reanimation, then the implementation is simple and follows current MPEG standards, but the mapping accuracy is unreliable especially for avatars with distinct geometry representations
Solution Approach 1:
The patent transitions from static pre-computed correspondences to dynamic per-vertex mapping where each vertex of the volumetric mesh independently maps to its corresponding polygon in the shadow mesh. This dynamic approach allows the mapping to adapt to different avatar geometries and animation states, improving accuracy while maintaining implementation feasibility through standardized data structures.
Solution Approach 2:
The patent changes the fundamental parameter of correspondence representation from polygon-level to vertex-level precision. By storing and processing individual vertex correspondences rather than bulk polygon mappings, the system achieves higher mapping accuracy for diverse avatar geometries while using established MPEG data structures to manage the increased data volume.
2Manufacturing precision
If vertex-level correspondence mapping is implemented for accurate avatar animation, then the mapping accuracy improves, but the data complexity and processing requirements increase
Solution Approach 1:
The patent creates a universal correspondence data structure that works across different avatar types and animation scenarios. The vertex-to-polygon mapping framework is designed to be format-agnostic and can handle various mesh topologies, making it applicable to diverse avatars without requiring format-specific implementations, thus managing complexity through standardization.
Solution Approach 2:
The patent introduces an intermediary correspondence data structure that bridges the shadow mesh and volumetric mesh representations. This intermediate layer stores vertex-to-polygon mappings that can be processed independently, decoupling the complexity of high-precision mapping from the animation rendering pipeline and making the system more manageable.
3Adaptability or versatility
If dynamic mesh sequences and timed-based updates are supported, then real-time avatar animation capability is enabled, but the computational load and processing time increase
Solution Approach 1:
The patent performs preliminary computation of vertex correspondences during avatar creation or initialization, storing these mappings in advance. During real-time animation, the pre-computed correspondences are reused and updated incrementally, avoiding the need to recalculate complex geometric relationships from scratch, thus reducing real-time processing time while maintaining dynamic animation capability.
Data Source
Figure 1
Figure 2~3
Figure 4~5
AI summary
Some embodiments of a method may include: obtaining information corresponding to an avatar, wherein the information comprises a first avatar representation associated with a parameter; mapping the first avatar representation to a second avatar representation, storing the second avatar representation in a data structure corresponding to the associated parameter; and streaming the second avatar representation to an external device.