T3HG-Editor: Text-driven 3D Human Garment Editing with Body Priors Embedded in SMPL-X

📄 arXiv: 2607.13654v1 📥 PDF

作者: Shaoru Sun, Xingtao Wang, Zihan Ma, Wenrui Li, Jiantao Zhou, Debin Zhao, Xiaopeng Fan

分类: cs.CV

发布日期: 2026-07-15

备注: 9 pages, 7 figures, 2 tables


💡 一句话要点

提出T3HG-Editor以解决文本驱动的3D人类服装编辑问题

🎯 匹配领域: 支柱六:视频提取与匹配 (Video Extraction)

关键词: 3D人类编辑 文本驱动 高斯编辑 SMPL-X 服装一致性 虚拟人类 计算机视觉

📋 核心要点

  1. 现有的3D人类服装编辑方法在保真度和一致性方面存在显著不足,导致编辑结果不理想。
  2. T3HG-Editor通过利用SMPL-X模型的几何和关节先验信息,提出了一种新的文本驱动编辑方法,确保高保真和服装一致性。
  3. 实验结果显示,T3HG-Editor在多个受试者和不同服装类型上均优于现有方法,提升了编辑质量和一致性。

📝 摘要(中文)

尽管3D高斯编辑(3DGE)取得了显著进展,但文本驱动的3D人类服装编辑仍然未得到充分探索。现有的3DGE方法通常将2D编辑技术应用于多视角渲染图像,并根据修改后的图像更新3D高斯。然而,这种方法在3D人类服装编辑中面临低保真度的问题,主要由于引入的失真和服装不一致性。为此,本文提出了一种名为T3HG-Editor的文本驱动3D人类服装编辑器,通过利用SMPL-X模型中嵌入的几何和关节先验信息,实现高保真和服装一致的结果。T3HG-Editor包含三个阶段:可编辑高斯的获取、服装一致性编辑和高斯更新与溢出修剪。实验表明,T3HG-Editor在编辑质量和服装一致性方面优于现有最先进的方法。

🔬 方法详解

问题定义:本文旨在解决文本驱动的3D人类服装编辑中的低保真度和服装不一致性问题。现有方法在多视角渲染图像上进行编辑,导致结果失真和服装不一致。

核心思路:T3HG-Editor通过利用SMPL-X模型中嵌入的几何和关节先验信息,设计了一个三阶段的编辑流程,以确保高保真和服装一致性。

技术框架:该方法分为三个主要阶段:首先获取可编辑的高斯;其次进行服装一致性编辑;最后通过溢出修剪更新高斯。每个阶段都针对特定问题进行优化。

关键创新:T3HG-Editor的主要创新在于通过SMPL-X模型的先验信息实现高保真度和服装一致性,而不需要额外的训练。这与传统方法的依赖于2D图像的方式形成鲜明对比。

关键设计:在获取可编辑高斯时,沿SMPL-X法线生成高斯,并通过2D掩码约束精确定位目标高斯。服装一致性编辑通过聚合多个视角的SMPL-X顶点对应的标记来实现,而高斯更新则利用定义在SMPL-X上的有符号距离函数(SDF)构建人类距离场,并结合2D语义掩码进行溢出修剪。

🖼️ 关键图片

fig_0
fig_1
fig_2

📊 实验亮点

实验结果表明,T3HG-Editor在多个受试者和不同服装类型上均优于现有最先进的方法,编辑质量和服装一致性显著提升,具体性能数据未详细列出,但整体效果明显优于基线方法。

🎯 应用场景

T3HG-Editor在虚拟人类服装设计、游戏角色定制和动画制作等领域具有广泛的应用潜力。其高保真和一致性的编辑能力能够提升虚拟服装的真实感和用户体验,推动相关行业的发展。未来,该技术可能会与其他AI驱动的设计工具结合,进一步拓展应用场景。

📄 摘要(原文)

While 3D Gaussian Editing (3DGE) has seen substantial progress, text-driven 3D human garment editing remains largely underexplored. Existing 3DGE works typically follow a paradigm that applies 2D editing techniques to multi-view rendered images and updates 3D Gaussians based on the modified images. Extending such methods to 3D human garment editing suffers from low-fidelity outcomes, caused by introduced distortions and garment inconsistencies. A promising breakthrough opportunity arises from the SMPL eXpressive (SMPL-X) model that embodies rich prior information for virtual humans. Motivated by this insight, we propose a text-driven 3D human garment editor termed T3HG-Editor, which delivers high-fidelity and garment consistent results by leveraging geometry and joint priors embedded in SMPL-X. Specifically, T3HG-Editor contains three stages, namely obtainment of editable Gaussians, garment consistent editing, and Gaussian updating with overflow pruning. The obtainment of editable Gaussians begins with seeding Gaussians along SMPL-X normals to generate sufficient near surface Gaussians, followed by a 2D mask constraint that precisely localizes the target Gaussians to be edited. The garment consistent editing aggregates tokens corresponding to the same SMPL-X vertex across multiple views and propagates them to their original views, enforcing garment consistency without requiring additional training. Gaussian updating with overflow pruning employs a Signed Distance Function (SDF) defined on SMPL-X to construct a human distance field, which is then integrated with a 2D semantic mask to prune overflowing Gaussians, thus preventing contamination of non-target regions. Experiments on multiple subjects and diverse garment types demonstrate that T3HG-Editor outperforms state-of-the-art methods in both editing quality and garment consistency.