Periodic Autoencoder Motion Alignment for Realistic Game Characters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for generating realistic character motion in electronic games are labor-intensive and lack the ability to capture the nuanced, asynchronous movements of real-world actors, requiring substantial manual tuning and limiting the realism of in-game character movements.
Innovation Solution
A system utilizing a periodic autoencoder to generate animation by analyzing motion capture information, employing local motion phase channels to automate the generation of realistic and nuanced character movements, allowing for real-time user input and enhanced temporal alignment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual tuning techniques are used to create realistic character motion, then the realism of character movement is improved, but the development time and complexity increase substantially
Solution Approach 1:
The patent replaces manual mechanical tuning of character animations with an automated neural network system. The neural network processes motion capture data and generates animations automatically, eliminating the need for designers to manually adjust skeleton positions while maintaining high realism through learned motion patterns from real human movement.
Solution Approach 2:
The system creates virtual copies of real human motion through motion capture technology. By recording actual human actors performing movements and using these captured motions as training data for the neural network, the system replicates realistic human motion patterns without requiring manual recreation by designers.
2Loss of time
If templates are used to automatically generate character motion, then development time is reduced, but the realism and nuance of movement are lost
Solution Approach 1:
The neural network dynamically adjusts motion parameters (position, orientation, timing) based on learned patterns from motion capture data rather than using fixed templates. This allows the system to generate varied, nuanced movements that adapt to different contexts while maintaining realism, overcoming the limitations of rigid template-based approaches.
3Reliability
If detailed character models and skeletons are created, then the realism of the in-game world is improved, but the complexity of movement implementation increases
Solution Approach 1:
The complex manual process of implementing detailed skeleton movements is replaced with a neural network that automatically processes motion capture data and generates appropriate animations. This substitution reduces the complexity of movement implementation while maintaining the realism provided by detailed character models.
Data Source
AI summary
The present disclosure provides a periodic autoencoder that can be used to generate a general motion manifold structure using local periodicity of the movement whose parameters are composed of phase, frequency, and amplitude. The periodic autoencoder is a novel neural network architecture that can learn periodic features from large unstructured motion datasets in an unsupervised manner. The character movements can be decomposed into multiple latent channels that can capture the non-linear periodicity of different body segments during synchronous, asynchronous, and transition movements while progressing forward in time, such that it captures spatial data and temporal data associated with the movements.