Analytically Integratable Zero-restlength Springs for Capturing Dynamic Modes unrepresented by Quasistatic Neural Networks

TL;DR

This paper introduces a hybrid modeling approach combining neural networks for quasistatic information with an analytically integrable spring model to efficiently capture dynamic modes in real-time simulations, reducing data needs and improving stability.

Contribution

The novel paradigm separates quasistatic and dynamic modeling, using neural networks for static info and simple springs for dynamic modes, enabling stable, data-efficient real-time simulation.

Findings

Effective capture of dynamic modes with simple spring models

Robust learning of spring parameters from limited data

Application demonstrated on soft-tissue dynamics in human bodies

Abstract

We present a novel paradigm for modeling certain types of dynamic simulation in real-time with the aid of neural networks. In order to significantly reduce the requirements on data (especially time-dependent data), as well as decrease generalization error, our approach utilizes a data-driven neural network only to capture quasistatic information (instead of dynamic or time-dependent information). Subsequently, we augment our quasistatic neural network (QNN) inference with a (real-time) dynamic simulation layer. Our key insight is that the dynamic modes lost when using a QNN approximation can be captured with a quite simple (and decoupled) zero-restlength spring model, which can be integrated analytically (as opposed to numerically) and thus has no time-step stability restrictions. Additionally, we demonstrate that the spring constitutive parameters can be robustly learned from a surprisingly small amount of dynamic simulation data. Although we illustrate the efficacy of our approach by considering soft-tissue dynamics on animated human bodies, the paradigm is extensible to many different simulation frameworks.