Avatar Animation Predictor Signaling With Hierarchical Dictionaries

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing solutions for signaling avatar animation parameters predictors are limited, as they do not account for the need to share neural networks trained on specific data formats, requiring additional preprocessing and postprocessing steps that are not easily communicated, leading to interoperability issues.

Innovation Solution

A hierarchical dictionary structure is used to signal avatar animation parameters predictors, incorporating a list of user inputs, processing steps, and functions, enabling seamless integration and prediction of animation parameters across various input types.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If a hierarchical dictionary structure is used to signal avatar animation parameters predictors, then interoperability and support for diverse input types are improved, but the device complexity and data structure complexity increase

Engineering Contradiction:
ImproveinteroperabilityVSAvoiddata structure complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The signaling structure is divided into hierarchical levels: a root dictionary containing high-level predictor information, and nested child dictionaries containing detailed processing step information. This segmentation allows the system to manage complexity by organizing data in manageable, structured units that can be processed independently while maintaining overall interoperability.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces a hierarchical dimension to the data structure, moving from a flat signaling approach to a multi-level nested structure. This adds an organizational dimension that enables better management of complex preprocessing and postprocessing steps while maintaining a clear hierarchy that facilitates interoperability across different input types and devices.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Adaptability or versatility

If preprocessing and postprocessing steps are encoded in the signaling structure, then support for specific data formats is improved, but the ease of operation and implementation difficulty increase

Engineering Contradiction:
Improvedata format supportVSAvoidimplementation difficulty
Core Design Contradiction:
Adaptability or versatilityVSEase of operation

Solution Approach 1:

The signaling structure encodes preprocessing and postprocessing steps in advance, before the actual avatar animation parameter prediction is performed. This preliminary encoding of required processing steps allows the system to automatically handle format-specific requirements without increasing operational complexity during runtime, as all necessary processing information is already prepared and structured in the signaling data.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4651088A1Avatar animation parameters predictor signaling
Publication Date: 2025.11.19 INTERDIGITAL CE PATENT HOLDINGS SAS
  • EP4651088A1 patent drawingFigure 1~2
  • EP4651088A1 patent drawingFigure 3
  • EP4651088A1 patent drawingFigure 4

AI summary

In one implementation, a user device may obtain an avatar model and a list of functions to be performed with associated parameters. For each function in the list of functions, the user device may create the function with associated parameters and execute the function to obtain processed data as input to next function in the list of functions when a current function is not a last function in the list of functions or as output parameters when the current function is the last function in the list of functions. Then the user device may animate the avatar model based on the output parameters. At the server side, for an avatar model, the server can generate the list of functions and associated parameters given a user input type and a prediction method, and transmit them to the user device.