Hybrid Neural-MPM for Interactive Fluid Simulations in Real-Time

2025-05-25

AI Summary

Traditional physics-based methods for real-time interactive fluid simulation suffer from high computational cost and latency, while existing machine learning approaches struggle to simultaneously achieve high fidelity and real-time performance (<33 ms). To address this, we propose Neural-Physics–MPM Co-Simulation: a framework integrating a differentiable Material Point Method (MPM) solver with a generative diffusion controller. We introduce the first hand-drawing-driven dynamic external force field generation via inverse-modeling training. Additionally, we design a real-time fallback mechanism between neural prediction and the classical solver to ensure stability and physical consistency. Evaluated on 2D/3D multi-material and obstacle-rich scenarios, our method achieves end-to-end latency of only 11–29% of baseline runtime, enabling >30 FPS interactive simulation and reducing physical error by 42%. Code and data are publicly available.

Technology Category

• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior
• Search and Optimization: Sampling/Simulation-based Search
• Computer Vision: Low Level & Physics-based Vision

Application Category

• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Web Mining and Content Analysis: Web data generation and simulation
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

We propose a neural physics system for real-time, interactive fluid simulations. Traditional physics-based methods, while accurate, are computationally intensive and suffer from latency issues. Recent machine-learning methods reduce computational costs while preserving fidelity; yet most still fail to satisfy the latency constraints for real-time use and lack support for interactive applications. To bridge this gap, we introduce a novel hybrid method that integrates numerical simulation, neural physics, and generative control. Our neural physics jointly pursues low-latency simulation and high physical fidelity by employing a fallback safeguard to classical numerical solvers. Furthermore, we develop a diffusion-based controller that is trained using a reverse modeling strategy to generate external dynamic force fields for fluid manipulation. Our system demonstrates robust performance across diverse 2D/3D scenarios, material types, and obstacle interactions, achieving real-time simulations at high frame rates (11~29% latency) while enabling fluid control guided by user-friendly freehand sketches. We present a significant step towards practical, controllable, and physically plausible fluid simulations for real-time interactive applications. We promise to release both models and data upon acceptance.

Problem

• Real-time interactive fluid simulation with low latency
• Balancing computational efficiency and physical fidelity
• Enabling user-guided fluid control via sketches

Innovation

• Hybrid neural and numerical simulation method
• Diffusion-based controller for fluid manipulation
• Fallback safeguard to classical numerical solvers