HairGPT: Strand-as-Language Autoregressive Modeling for Realistic 3D Hairstyle Synthesis

2026-05-09

AI Summary

Existing 3D hairstyle generation methods struggle to simultaneously capture global structure and local details while lacking semantic controllability, resulting in limited realism and editing flexibility. This work introduces a novel approach by modeling hair strands as language-like autoregressive sequences and proposes a spatially and structurally disentangled generative framework. It employs a geometric tokenizer to construct hierarchical strand representations and integrates region-aware semantic annotations to guide progressive synthesis under scalp-layout constraints. Additionally, a digital combing alignment mechanism is introduced to enhance geometric consistency. The method enables high-fidelity, semantically controllable 3D hairstyle generation with flexible editability and strong generalization to rare and stylized hairstyles, significantly outperforming existing approaches in both realism and creative freedom.

Abstract

Hair is a rich medium of visual and cultural expression, yet its digital modeling remains challenging due to the duality of fluidity and structure. Many existing generative approaches rely primarily on continuous diffusion fields, which entangle global topology with local texture and obscure the semantic and structural organization of hairstyles. To address this, we propose HairGPT, a strand-centric framework that treats strands as generative primitives and formulates realistic 3D hairstyle synthesis as a dual-decoupled autoregressive sequence modeling problem. Our method applies spatial decoupling across semantic scalp regions and structural decoupling along a hierarchical strand representation, progressing from global layout to fine-grained style. We further introduce a geometric tokenizer and region-aware semantic annotations to guide strand-level generation, enabling compositional editing, synthesis of rare and complex hairstyles, and adaptation to stylized domains. By aligning generative modeling with the workflow of digital grooming, HairGPT turns hair generation from opaque texture synthesis into a structured and semantically controllable authoring process, supporting robust semantic conditioning and high-fidelity results across realistic and stylized domains. Project Page: https://haiminluo.github.io/hairgpt/

Problem

• 3D hairstyle synthesis
• strand modeling
• semantic structure
• autoregressive generation
• hair geometry

Innovation

• strand-centric modeling
• autoregressive sequence modeling
• spatial-structural decoupling
• geometric tokenizer
• semantic conditioning