HairGS: Hair Strand Reconstruction based on 3D Gaussian Splatting

2025-09-09

AI Summary

Three-dimensional reconstruction of human hair is critical for virtual reality and digital human modeling; however, existing methods predominantly prioritize geometric fidelity while neglecting inter-hair connectivity and topological structure. This paper proposes the first hair-strand-level reconstruction framework based on 3D Gaussian splatting (3DGS), operating in two stages: initial geometric reconstruction from multi-view images, followed by topology-aware refinement. We introduce a novel strand-level topological evaluation metric and integrate Gaussian segment merging with photometrically supervised growth optimization to explicitly model and preserve hair connectivity. Technically, our approach unifies differentiable Gaussian rasterization, multi-stage geometric optimization, and photometric consistency constraints. Extensive experiments on both synthetic and real-world datasets demonstrate that our method achieves superior geometric fidelity and topological plausibility simultaneously, with most reconstructions completed within one hour—significantly outperforming state-of-the-art approaches.

Technology Category

• Computer Vision: 3D Computer Vision
• Humans and AI: Game Design — Virtual Humans, NPCs and Autonomous Characters
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG

Abstract

Human hair reconstruction is a challenging problem in computer vision, with growing importance for applications in virtual reality and digital human modeling. Recent advances in 3D Gaussians Splatting (3DGS) provide efficient and explicit scene representations that naturally align with the structure of hair strands. In this work, we extend the 3DGS framework to enable strand-level hair geometry reconstruction from multi-view images. Our multi-stage pipeline first reconstructs detailed hair geometry using a differentiable Gaussian rasterizer, then merges individual Gaussian segments into coherent strands through a novel merging scheme, and finally refines and grows the strands under photometric supervision. While existing methods typically evaluate reconstruction quality at the geometric level, they often neglect the connectivity and topology of hair strands. To address this, we propose a new evaluation metric that serves as a proxy for assessing topological accuracy in strand reconstruction. Extensive experiments on both synthetic and real-world datasets demonstrate that our method robustly handles a wide range of hairstyles and achieves efficient reconstruction, typically completing within one hour. The project page can be found at: https://yimin-pan.github.io/hair-gs/

Problem

• Reconstructing strand-level hair geometry from multi-view images
• Addressing connectivity and topology neglect in hair reconstruction
• Developing a new evaluation metric for topological accuracy

Innovation

• Extends 3DGS framework for strand-level hair reconstruction
• Uses differentiable Gaussian rasterizer and merging scheme
• Refines strands under photometric supervision