3D Avatar Eye and Eyelid Transforms for Morph Target Alignment

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current scene description formats fail to ensure accurate placement and rendering of eyes and eyelids when avatar shapes change, particularly due to the use of morph targets, leading to issues like overlapping or misalignment.

Innovation Solution

A method and device for encoding and transmitting scene descriptions that include eyelid and eye transform processes, using corrective scale and translation factors, animation-based transformations, and eyelid handling techniques to maintain correct positioning and orientation of eyes and eyelids during avatar shape changes.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If morph targets are used to change avatar shape, then avatar adaptability is improved, but eye and eyelid positioning accuracy deteriorates

Engineering Contradiction:
Improveavatar shape change capabilityVSAvoideye and eyelid positioning accuracy
Core Design Contradiction:
Adaptability or versatilityVSManufacturing precision

Solution Approach 1:

The patent applies preliminary action by pre-defining transform types and transform parameters for eye nodes, and eyelid handling types with handling parameters, in the scene description before rendering. These preliminary transformations ensure that when morph targets change the avatar shape, the eyes and eyelids are automatically repositioned according to the pre-specified transform rules, preventing misalignment and overlapping issues.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent utilizes parameter changes by modifying transform parameters (such as translation and rotation values) and handling parameters based on the morph target weight. When the avatar shape changes via morph targets, the system dynamically adjusts these parameters to maintain correct eye and eyelid positioning, thereby resolving the positioning accuracy issue while preserving shape adaptability.

Inventive Principle:
Principle #35Parameter changes

2Manufacturing precision

If eye and eyelid transformations are applied dynamically, then rendering accuracy is improved, but processing complexity increases

Engineering Contradiction:
Improverendering accuracyVSAvoidtransformation process complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent applies segmentation by dividing the transformation process into distinct, manageable components: eye node transformations (with separate transform types and parameters) and eyelid transformations (with separate handling types and parameters). This segmentation allows each component to be processed independently according to its specific rules, improving rendering accuracy while keeping the overall system organized and manageable despite the increased processing requirements.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP4679845A1Signaling avatar eye and eyelids handling in scene description
Publication Date: 2026.01.14 INTERDIGITAL CE PATENT HOLDINGS SAS
  • EP4679845A1 patent drawingFigure 1
  • EP4679845A1 patent drawingFigure 2~4
  • EP4679845A1 patent drawingFigure 5

AI summary

Methods, apparatus and data stream are provided for rendering a 3D scene comprising avatars having eyes associated with an eyelid The method comprises obtaining a scene description of the 3D scene comprising a list of eyelid descriptions. Every eyelid description comprises a reference to an eye node, a transform type for the eye node with associated transform parameters, and an eyelid handling type with associated handling parameters. Then the 3D scene is prepared for rendering according to the scene description. For each eye node of the avatars, if the avatar has changed, the eye and the eyelid may be transformed according to the eyelid description corresponding to the eye node. An eye transform process may be applied according to the transform type and transform parameters; and an eyelid transform process may be applied according to the eyelid handling type and handling parameters. Then, the 3D scene is rendered.