HyperBones: Realtime Bone-driven Neural Garment Simulation with Hypernetwork Conditioning

2026-05-19

AI Summary

Existing real-time cloth simulation methods struggle to simultaneously achieve complex dynamics for loose garments, computational efficiency, and physical realism. This work proposes a neural simulation approach driven by virtual bones, which jointly leverages a coarse-grained virtual bone network and a fine-grained convolutional neural mapping to decouple identity-dependent deformations while enabling efficient inference. The method innovatively integrates a dual-level neural architecture, a hypernetwork-based conditioning mechanism, and a physics-supervised training strategy that operates without external simulators. Evaluated on consumer-grade GPUs, the approach achieves over 300 frames per second and realistically simulates a variety of loose-fitting garments, demonstrating strong generalization across diverse body shapes and motion sequences.

Abstract

Recent advances in garment simulation have brought high-quality results closer to real-time performance. Physics-based simulators can produce accurate motion, but remain too computationally expensive for interactive applications. In contrast, linear blend skinning is efficient, but cannot capture the complex dynamics of loose-fitting garments, often leading to unrealistic motion and visual artifacts. Neural methods offer a promising alternative, yet they still struggle to animate loose clothing plausibly under strict runtime constraints. We present a fast and physically plausible approach for dynamic garment simulation. Our method trains a reduced-space neural dynamics simulator composed of independent coarse- and fine-level components. At the coarse level, the garment is driven by a set of virtual bones integrated with a lightweight neural network. Fine-scale wrinkle details are then recovered using a trained convolutional neural map. By decoupling identity-specific computation from real-time neural integration, our architecture maintains high performance while supporting diverse body shapes and motions. We further introduce an effective physics-supervision scheme that enables accurate results without relying on an external simulator. Experiments show that our method produces physically plausible garment dynamics, generalizes across a range of motions and body shapes, and supports a fixed set of garments. Our simulator runs at 300+ FPS on a commodity GPU, making it suitable for real-time applications.

Problem

• garment simulation
• real-time animation
• loose clothing
• neural dynamics
• physical plausibility

Innovation

• neural garment simulation
• hypernetwork conditioning
• bone-driven animation
• real-time performance
• physics-supervised learning