STAvatar：软绑定与时序密度控制的单目三维头部化身重建
STAvatar: Soft Binding and Temporal Density Control for Monocular 3D Head Avatars Reconstruction
把高斯绑定从硬性网格三角形改为 UV 空间的可学习软偏移，使高斯泼溅表示仍能接受自适应密度控制，这一思路可直接迁移到人体、手部等其它基于网格模板的可动高斯泼溅重建中；其融合感知误差的克隆准则也为高斯致密化提供了可复用的判据。
原文依据 a UV-Adaptive Soft Binding framework that leverages both image-based and geometric priors to learn per-Gaussian feature offsets within the UV space
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论从单目视频重建高保真且可驱动的三维头部化身仍是一项具有挑战性但至关重要的任务。现有基于三维高斯泼溅的方法通常将高斯绑定到网格三角形上，并仅通过线性混合蒙皮建模形变，导致运动僵硬且表达能力受限。此外，它们缺乏针对频繁被遮挡区域（如口腔内部、眼睑）的专门处理策略。为解决上述局限，我们提出 STAvatar，它包含两个关键组件：（1）UV 自适应软绑定框架，同时利用基于图像的先验与几何先验，在 UV 空间中学习每个高斯的高斯特征偏移。该 UV 表示支持动态重采样，从而与自适应密度控制（ADC）完全兼容，并增强了对形状与纹理变化的适应能力。（2）时序 ADC 策略，首先对结构相似的帧进行聚类，以便更有针对性地计算致密化判据；进而引入一种新的融合感知误差作为克隆判据，联合捕捉几何与纹理差异，促使在需要更精细细节的区域进行致密化。在四个基准数据集上的大量实验表明，STAvatar 取得了最优的重建性能，尤其在捕捉细粒度细节和重建频繁被遮挡区域方面表现突出。项目主页见 https://github.com/JiankuoZhao/STAvatar.git。
英文原文摘要 1498 字符
Reconstructing high-fidelity and animatable 3D head avatars from monocular videos remains a challenging yet essential task. Existing methods based on 3D Gaussian Splatting typically bind Gaussians to mesh triangles and model deformations solely via Linear Blend Skinning, which results in rigid motion and limited expressiveness. Moreover, they lack specialized strategies to handle frequently occluded regions (e.g., mouth interiors, eyelids). To address these limitations, we propose STAvatar, which consists of two key components: (1) a UV-Adaptive Soft Binding framework that leverages both image-based and geometric priors to learn per-Gaussian feature offsets within the UV space. This UV representation supports dynamic resampling, ensuring full compatibility with Adaptive Density Control (ADC) and enhanced adaptability to shape and textural variations. (2) a Temporal ADC strategy, which first clusters structurally similar frames to facilitate more targeted computation of the densification criterion. It further introduces a novel fused perceptual error as clone criterion to jointly capture geometric and textural discrepancies, encouraging densification in regions requiring finer details. Extensive experiments on four benchmark datasets demonstrate that STAvatar achieves state-of-the-art reconstruction performance, especially in capturing fine-grained details and reconstructing frequently occluded regions. Our project is available at https://github.com/JiankuoZhao/STAvatar.git.