Avatar Asset Resource Optimization via Blend Shape Reuse

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Mobile devices face challenges in providing a wide range of avatar customization options without consuming excessive memory and processing resources, leading to delays and a poor user experience due to the need to store and load numerous customization assets and blend shapes.

Innovation Solution

The method involves optimizing avatar asset resources by storing only the most frequently used features and assets, using a universal rig for the head, and reusing blend shapes across multiple assets, with a simulation engine determining linear combinations of blend shapes for animations, and employing model swapping for hairstyles to reduce the number of required blend shapes.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If all customization options are stored and made available continuously, then users have access to a wide variety of avatar customization options, but memory usage and processing resources increase significantly

Engineering Contradiction:
Improveavatar customization optionsVSAvoidmemory usage
Core Design Contradiction:
Adaptability or versatilityVSQuantity of substance

Solution Approach 1:

The patent segments avatar customization options into frequently used and less frequently used categories. Only frequently used features and assets are stored in memory, while other options are loaded on-demand or generated procedurally. This segmentation allows the system to provide diverse customization options without storing all assets simultaneously, thus reducing memory usage while maintaining versatility.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements a universal rig system that can be reused across multiple avatar assets and customization options. Instead of creating separate rigs for each asset, a single universal rig serves multiple functions, reducing the overall quantity of data that needs to be stored and processed while still providing diverse customization capabilities.

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

2Adaptability or versatility

If all customization options are stored and made available continuously, then users have access to comprehensive avatar options, but processing time and device performance are degraded

Engineering Contradiction:
Improveavatar customization optionsVSAvoidprocessing speed
Core Design Contradiction:
Adaptability or versatilityVSProductivity

Solution Approach 1:

The system segments customization options and loads only the necessary portions into memory based on user selection. Frequently used features are pre-loaded, while other options are loaded on-demand. This approach maintains comprehensive customization capabilities while significantly reducing the amount of data that needs to be processed simultaneously, thus improving processing speed and device productivity.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements preliminary action by pre-loading and caching frequently used avatar features and assets in memory before they are needed. This allows the system to provide fast access to commonly selected customization options without requiring full asset loading each time, thereby improving processing speed while maintaining comprehensive options.

Inventive Principle:
Principle #10Preliminary action

3Adaptability or versatility

If all customization options are loaded and displayed, then users can see all available avatar options, but loading time and user experience are negatively impacted

Engineering Contradiction:
Improveavatar customization optionsVSAvoidloading time
Core Design Contradiction:
Adaptability or versatilityVSLoss of time

Solution Approach 1:

The patent segments avatar customization options into frequently used and less frequently used categories. Only frequently used features and assets are pre-loaded into memory, while other options are loaded on-demand when selected by the user. This segmentation dramatically reduces initial loading time while still providing access to comprehensive customization options, thus improving user experience without sacrificing versatility.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system performs preliminary action by pre-loading and caching frequently selected avatar features, assets, and blend shapes into memory before the user needs them. This allows the application to start quickly and provide instant access to commonly used customization options, reducing perceived loading time while maintaining comprehensive customization capabilities.

Inventive Principle:
Principle #10Preliminary action

4Reliability

If multiple blend shapes are created for each asset, then detailed avatar animations and expressions are achieved, but memory usage and processing resources increase

Engineering Contradiction:
Improveanimation qualityVSAvoidmemory usage
Core Design Contradiction:
ReliabilityVSQuantity of substance

Solution Approach 1:

The patent implements a universal blend shape system where a single set of blend shapes can be applied to multiple different assets through the universal rig. Instead of creating separate blend shapes for each asset, the same blend shapes are reused across multiple assets, maintaining detailed animation and expression quality while significantly reducing memory usage by eliminating redundant data.

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

Data Source

PatentEP4123587A1Optimized avatar asset resource
Publication Date: 2023.01.25 APPLE INC
  • EP4123587A1 patent drawingFigure 1A~1B
  • EP4123587A1 patent drawingFigure 1C~1D
  • EP4123587A1 patent drawingFigure 1E~1F

AI summary

Various customization options are provided for customizing a 3D avatar of a head. Features of the head and assets corresponding to the features can be customized using blend shapes. An amount of storage for the plurality of blend shapes is minimized by determining overlapping blend shapes that can be reused for a plurality of different assets. Further, techniques are provided for dynamic changes to an avatar in accordance with selected features and assets.