基于频域去噪先验的遮挡人体运动捕捉
Occluded Human Body Capture with Frequency Domain Denoising Prior
把遮挡下的单目人体运动捕捉转化为小波系数选择问题，用频域扩散模型学习长期周期性与动量先验来修正被遮挡 2D 关键点的概率分布，这一「频域去噪先验」思路可迁移到其他存在长期缺失观测的时序视觉任务（如遮挡跟踪、时序姿态补全）。
原文依据 we exploit reliable image observations in frequency domain and formulate the motion capture task as a wavelet coefficients selection process
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论单目遮挡场景下的人体运动捕捉面临显著挑战。尽管已有少数工作显式考虑了遮挡问题，但基于图像的方法因缺乏时序约束而不可靠，基于视频的方法又无法从时域运动先验中获得足够知识以应对长期遮挡。然而，被遮挡的人体运动通常呈现周期性模式与一致的动量。受此观察启发，我们利用频域中可靠的图像观测，将运动捕捉任务建模为小波系数选择过程。具体而言，我们首先为被遮挡的 2D 关键点构建概率分布，随后引入频域扩散模型，借助离散小波变换（DWT）学习长期周期信息与物理动量来精化该分布。由此得到的去噪先验能够筛选有效的小波分量，以配合 3D 解码器完成 3D 运动捕捉。通过采用联合重投影策略，我们还能用同一扩散过程训练 3D 解码器。为进一步推动人体遮挡相关任务，我们还提出了首个 3D 遮挡运动数据集 OcMotion，它可作为训练与评估的新基准。实验结果表明，我们的方法能够从遮挡视频中生成准确且连贯的人体运动。更多信息见 https://github.com/boycehbz/FreqMotion。
英文原文摘要 1518 字符
Monocular human motion capture in occlusion scenarios presents significant challenges. Although a few works have explicitly considered the occlusion problem, image-based methods are unreliable due to the lack of temporal constraints while video-based approaches cannot gain sufficient knowledge from time domain motion priors to address long-term occlusions. However, occluded human motion typically exhibits periodic patterns and consistent momentum. Inspired by this observation, we exploit reliable image observations in frequency domain and formulate the motion capture task as a wavelet coefficients selection process. Specifically, we first construct probabilistic distributions for the occluded 2D keypoints, and then introduce a frequency domain diffusion model to refine the distributions by learning long-term periodic information and physical momentum with Discrete Wavelet Transform (DWT). Consequently, the learned denoising prior can select valid wavelet components to facilitate the 3D motion capture with a 3D decoder. By employing a joint reprojection strategy, we can also use the same diffusion process to train the 3D decoder. To further promote human occlusion-related tasks, we also present the first 3D occluded motion dataset, OcMotion, which serves as a new benchmark for both training and evaluation. Experimental results demonstrate that our method can produce accurate and coherent human motions from occluded videos. More information is available at https://github.com/boycehbz/FreqMotion.