Clothing Deformation Model for Realistic Garment Simulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current physically based simulation (PBS) methods for clothing are inflexible and require high-resolution meshes, complex functions, and time-consuming collision handling, limiting their ability to generalize to new body shapes and poses, making them inefficient for simulating realistic clothing on diverse human forms.
Innovation Solution
A system and method for generating a clothing deformation model that factors pose-independent and pose-dependent deformations, allowing for the automatic creation of custom-shaped garments that can be adapted to various body shapes and poses, enabling realistic clothing simulation without manual intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If physically based simulation (PBS) is used to model clothing behavior, then realism of clothing simulation is improved, but computational time and complexity increase significantly
Solution Approach 1:
The system performs preliminary actions by pre-computing clothing simulations for a comprehensive set of body shapes and poses during an offline training phase. These pre-computed results are stored in a database, allowing the system to quickly retrieve and interpolate results for new inputs without performing time-consuming PBS simulations in real-time, thus resolving the contradiction between realism and computational time
Solution Approach 2:
The system creates simplified copy representations of complex PBS simulation results by training machine learning models (such as autoencoders or regression models) on pre-computed simulation data. These models capture the essential clothing behavior patterns and can generate realistic clothing appearances much faster than original PBS simulations, maintaining realism while dramatically reducing computational time
2Reliability
If high resolution meshes are used to represent clothing details, then realism of folds and wrinkles is improved, but device complexity and computational requirements increase
Solution Approach 1:
The system replaces the mechanical PBS simulation approach with a data-driven machine learning model. Instead of using complex high-resolution meshes and physics calculations, the trained model directly predicts clothing appearance from body inputs, substituting the mechanical simulation system with an intelligent system that achieves similar or better detail representation with lower computational complexity
3Manufacturing precision
If separate PBS simulations are performed for every body shape, then accuracy for specific body models is improved, but productivity and efficiency decrease
Solution Approach 1:
The system develops a universal machine learning model that can handle multiple body shapes and poses simultaneously. The model is trained on diverse body data and can generalize to new body types without requiring separate simulations, making the system multi-functional and significantly improving productivity when dressing multiple bodies with different shapes
Solution Approach 2:
The system performs preliminary training on a comprehensive dataset of various body shapes and poses before deployment. This pre-computation allows the model to accurately handle any body type within the training distribution, eliminating the need for separate simulations for each body shape while maintaining high accuracy
4Reliability
If complicated non-linear functions are used to capture fabric properties, then realism of fabric behavior is improved, but ease of manufacture and implementation decrease
Solution Approach 1:
The system replaces complicated non-linear physics-based functions with a data-driven machine learning model that learns fabric behavior patterns from training data. The model captures complex fabric properties through learned representations rather than explicit physics equations, making the system easier to implement and modify while maintaining realistic fabric behavior
Solution Approach 2:
The system changes the approach from using fixed physics parameters to learning parameters from data. Instead of manually specifying fabric properties in complex non-linear functions, the model learns appropriate parameters and relationships from training examples, simplifying implementation while preserving accuracy
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Systems, methods, and computer-readable storage media for simulating realistic clothing. The system generates a clothing deformation model for a clothing type, wherein the clothing deformation model factors a change of clothing shape due to rigid limb rotation, pose-independent body shape, and pose-dependent deformations. Next, the system generates a custom-shaped garment for a given body by mapping, via the clothing deformation model, body shape parameters to clothing shape parameters. The system then automatically dresses the given body with the custom- shaped garment.