Self-supervised Garment Dynamics with Persistent Wrinkles

2026-06-23

AI Summary

Existing self-supervised neural cloth simulation methods struggle to model permanent wrinkles caused by fabric plasticity, compromising visual realism. This work addresses this limitation by explicitly incorporating elastoplastic mechanics into a self-supervised framework for the first time. We propose a dynamic energy minimization loss function and employ a curriculum learning strategy that progressively transitions from purely elastic to elastoplastic material behavior, enabling self-supervised learning of persistent wrinkles. Our approach generates natural and visually plausible wrinkles across diverse garment types, body shapes, and motions, outperforming current methods on key metrics and significantly enhancing the visual fidelity of simulated cloth.

Abstract

The self-supervised neural garment simulator has become popular due to its high efficiency, good visual realism, and no reliance on training data. However, existing methods greatly simplify the mechanical properties of fabrics, ignoring persistent wrinkles caused by plasticity. Although this simplification allows for modeling of purely elastic material and simple training via energy minimization, the lack of believable wrinkles adversely affects the visual realism. Therefore, we introduce the first self-supervised neural garment simulator that explicitly models persistent wrinkles. This is achieved by a novel physics-inspired loss function, turning learning into a moving energy minimization problem to mimic plasticity. However, this requires learning to use a changing loss function, which causes difficulties in training i.e. the loss function changes during training. To this end, we propose a new physics-inspired curriculum learning scheme where the target material for learning gradually changes from pure elasticity to elasto-plasticity, allowing the loss function and the learnable parameters to jointly converge. Through a comprehensive evaluation, we show that for the first time, self-supervised learning models can generate natural persistent wrinkles, outperforming existing methods on a variety of garments, body shapes, and body motions, according to a range of metrics.

Problem

• self-supervised learning
• garment simulation
• persistent wrinkles
• fabric plasticity
• visual realism

Innovation

• self-supervised learning
• garment simulation
• persistent wrinkles
• elasto-plasticity
• curriculum learning