UIKA：基于无位姿图像的快速通用头部化身
UIKA: Fast Universal Head Avatar from Pose-Free Images
用像素级人脸对应把任意视角、任意表情的输入重投影到与相机位姿和表情无关的 UV 空间，再在屏幕级与 UV 级联合做注意力，为无位姿输入的前馈可动画头部重建提供了可复用的表示与网络设计思路。
原文依据 we introduce a UV-guided avatar modeling strategy, in which each input image is associated with a pixel-wise facial correspondence estimation
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论我们提出 UIKA，一种可从任意数量无位姿输入（包括单张图像、多视角采集和智能手机拍摄的视频）进行前馈式可动画高斯头部建模的方法。与传统化身方法不同——后者需要影棚级多视角采集系统，并通过长时间优化过程重建特定于某个人的模型——我们从模型表示、网络设计和数据准备三个角度重新审视该任务。首先，我们引入一种 UV 引导的化身建模策略，其中每张输入图像都与一个逐像素的人脸对应估计相关联。这种对应估计使我们能够将每个有效像素颜色从屏幕空间重投影到 UV 空间，而该空间与相机位姿和人物表情无关。此外，我们设计了可学习的 UV 词元，注意力机制可在屏幕级和 UV 级两个层面上作用于这些词元。学习到的 UV 词元可利用来自所有输入视角的聚合 UV 信息，被解码为规范高斯属性。为训练我们的大规模化身模型，我们还额外准备了一个大规模、身份多样的合成训练数据集。我们的方法在单目和多视角设置下均显著优于现有方法。更多细节见项目主页：https://zijian-wu.github.io/uika-page/。
英文原文摘要 1328 字符
We present UIKA, a feed-forward animatable Gaussian head model from an arbitrary number of pose-free inputs, including a single image, multi-view captures, and smartphone-captured videos. Unlike the traditional avatar method, which requires a studio-level multi-view capture system and reconstructs a human-specific model through a long-time optimization process, we rethink the task through the lenses of model representation, network design, and data preparation. First, we introduce a UV-guided avatar modeling strategy, in which each input image is associated with a pixel-wise facial correspondence estimation. Such correspondence estimation allows us to reproject each valid pixel color from screen space to UV space, which is independent of camera pose and character expression. Furthermore, we design learnable UV tokens on which the attention mechanism can be applied at both the screen and UV levels. The learned UV tokens can be decoded into canonical Gaussian attributes using aggregated UV information from all input views. To train our large avatar model, we additionally prepare a large-scale, identity-rich synthetic training dataset. Our method significantly outperforms existing approaches in both monocular and multi-view settings. See more details in our project page: https://zijian-wu.github.io/uika-page/.