HairFormer: Transformer-Based Dynamic Neural Hair Simulation

TL;DR

HairFormer introduces a Transformer-based neural framework that achieves real-time, high-fidelity dynamic hair simulation across diverse styles and motions, effectively handling complex and unseen hairstyles with broad generalization.

Contribution

This work is the first to utilize Transformer architectures for dynamic hair simulation, enabling broad generalization and real-time performance for complex hairstyles and motions.

Findings

Achieves real-time inference for static and dynamic hair.

Handles complex, unseen hairstyles with high fidelity.

Uses physics-informed losses for realistic results.

Abstract

Simulating hair dynamics that generalize across arbitrary hairstyles, body shapes, and motions is a critical challenge. Our novel two-stage neural solution is the first to leverage Transformer-based architectures for such a broad generalization. We propose a Transformer-powered static network that predicts static draped shapes for any hairstyle, effectively resolving hair-body penetrations and preserving hair fidelity. Subsequently, a dynamic network with a novel cross-attention mechanism fuses static hair features with kinematic input to generate expressive dynamics and complex secondary motions. This dynamic network also allows for efficient fine-tuning of challenging motion sequences, such as abrupt head movements. Our method offers real-time inference for both static single-frame drapes and dynamic drapes over pose sequences. Our method demonstrates high-fidelity and generalizable dynamic hair across various styles, guided by physics-informed losses, and can resolve penetrations even for complex, unseen long hairstyles, highlighting its broad generalization.