4DEquine：解耦运动与外观的单目视频四维马科动物重建
4DEquine: Disentangling Motion and Appearance for 4D Equine Reconstruction from Monocular Video
将4D动物重建解耦为动态运动回归与静态外观重建，用时空Transformer从单目视频回归像素对齐的SMPL类序列，并用前馈网络从单张图像重建可动画的3D高斯化身，为单目视频4D重建提供了可复用的解耦范式与合成数据训练策略。
原文依据 we propose a novel framework called 4DEquine by disentangling the 4D reconstruction problem into two sub-problems: dynamic motion reconstruction and static appearance reconstruction.
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马科动物（如马）的四维结构对动物福利具有重要意义。以往主流的四维动物重建方法需要对整个视频中的运动与外观进行联合优化，耗时且对不完整观测敏感。本文提出一种新颖框架4DEquine，将四维重建问题解耦为两个子问题：动态运动重建与静态外观重建。在运动方面，引入一种简单而有效的时空Transformer，并配合后优化阶段，从视频中回归平滑且像素对齐的姿态与形状序列。在外观方面，设计一种新颖的前馈网络，仅需单张图像即可重建高保真、可动画的三维高斯化身。为辅助训练，构建大规模合成运动数据集VarenPoser，具有高质量表面运动与多样相机轨迹；以及合成外观数据集VarenTex，包含通过多视角扩散生成的真实感多视角图像。尽管仅在合成数据集上训练，4DEquine在真实世界APT36K与AiM数据集上取得最优性能，证明了4DEquine及新数据集在几何与外观重建上的优越性。全面的消融实验验证了运动与外观重建网络的有效性。代码与数据将公开。
英文原文摘要 1500 字符
4D reconstruction of equine family (e.g. horses) from monocular video is important for animal welfare. Previous mainstream 4D animal reconstruction methods require joint optimization of motion and appearance over a whole video, which is time-consuming and sensitive to incomplete observation. In this work, we propose a novel framework called 4DEquine by disentangling the 4D reconstruction problem into two sub-problems: dynamic motion reconstruction and static appearance reconstruction. For motion, we introduce a simple yet effective spatio-temporal transformer with a post-optimization stage to regress smooth and pixel-aligned pose and shape sequences from video. For appearance, we design a novel feed-forward network that reconstructs a high-fidelity, animatable 3D Gaussian avatar from as few as a single image. To assist training, we create a large-scale synthetic motion dataset, VarenPoser, which features high-quality surface motions and diverse camera trajectories, as well as a synthetic appearance dataset, VarenTex, comprising realistic multi-view images generated through multi-view diffusion. While training only on synthetic datasets, 4DEquine achieves state-of-the-art performance on real-world APT36K and AiM datasets, demonstrating the superiority of 4DEquine and our new datasets for both geometry and appearance reconstruction. Comprehensive ablation studies validate the effectiveness of both the motion and appearance reconstruction network. Code and data will be released.