多视图一致的三维高斯头部头像“无需”多视图生成
Multi-view Consistent 3D Gaussian Head Avatars 'without' Multi-view Generation
提出在 3D 高斯表示上直接施加多视图一致性约束的单次状态空间模型，无需多视图数据或中间视图合成，为单视图 3D 头像生成提供了可复用的 HiSS 块与 HiBiSS 扫描机制，以及基于 SE(3) 多视图评论器的自监督训练思路。
原文依据 we learn both conditional and unconditional 3D head models from randomly sampled 2D images alone, without using multi-view data, 3D supervision, or intermediate view generation.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论高保真三维高斯头部头像生成对于 AR/VR、远程呈现和数字人等应用至关重要。现有方法依赖多视图数据集、三维捕捉或中间二维视图合成。与之相反，我们仅从随机采样的二维图像中学习条件式和非条件式三维头部模型，不使用多视图数据、三维监督或中间视图生成。我们提出 MVCHead，一种单次状态空间模型，直接在三维表示中强制多视图一致性（MVC），并在这些约束下回归三维高斯。其核心是我们提出的分层状态空间（HiSS）块，它从粗到细逐步细化高斯，同时捕获长程依赖关系。在每个 HiSS 块内，我们用所提出的分层双向状态扫描（HiBiSS）修改 Mamba 的标准单向扫描，使递归与多视图不一致性最强的轴对齐。最后，我们设计了一个 SE(3) 多视图评论器，用于判断一组自渲染图像是否源自单一底层三维配置，从而在未观察真实多视图对的情况下奖励跨视图像素对齐。MVCHead 实现了最优的感知质量，在纹理和几何一致性上超越先前方法，并保持可比的形状一致性。为展示可扩展性，我们发布了 FaceGS-10K，这是首个用于训练和评估三维头部模型的大规模即用型三维高斯头部资产数据集。项目页面和代码：https://humansensinglab.github.io/MVCHead/
英文原文摘要 1605 字符
High-fidelity 3D Gaussian head avatar generation is critical for applications such as AR/VR, telepresence, and digital humans. Existing methods depend on multi-view datasets, 3D captures, or intermediate 2D view synthesis. In contrast, we learn both conditional and unconditional 3D head models from randomly sampled 2D images alone, without using multi-view data, 3D supervision, or intermediate view generation. We introduce MVCHead, a single-shot state space model that enforces multi-view consistency (MVC) directly in the 3D representation while regressing 3D Gaussians under these constraints. At its core, we propose a Hierarchical State Space (HiSS) block that progressively refines Gaussians from coarse to fine, while capturing long-range dependencies. Within each HiSS block, we modify Mamba's standard unidirectional scan with the proposed Hierarchical Bi-directional State Scan (HiBiSS) that aligns recurrence with the axes along which multi-view inconsistencies are strongest. Finally, we design an SE(3) Multi-view Critic that judges whether a set of self-renders arises from a single underlying 3D configuration, rewarding cross-view pixel alignment without observing real multi-view pairs. MVCHead achieves state-of-the-art perceptual quality, surpasses prior methods in both texture and geometric consistency, and maintains comparable shape consistency. To demonstrate scalability, we release FaceGS-10K, the first large-scale dataset of ready-to-use 3D Gaussian head assets for training and evaluation of 3D head models. Project Page and code: https://humansensinglab.github.io/MVCHead/