PoseD-Flow：通用且可引导的人体姿态流匹配模型
PoseD-Flow: Versatile and Guided Flow Matching Model of Human Pose
把流匹配先验直接定义在关节旋转的乘积流形上，并用可微 ODE 采样做免训练引导，可直接用于遮挡/噪声下的姿态补全、去噪与逆运动学求解，也为其他非欧几里得结构化输出（如手部、SMPL 参数）提供可复用的生成先验范式。
原文依据 Riemannian D-Flow, a principled guidance mechanism that, by differentiating through its ODE sampling dynamics, conditions PoseRFM at inference without any task-specific training
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论生成式姿态先验近来已成为在遮挡或噪声条件下进行推理的有力工具。然而，当前最强的生成范式——流匹配——由于两个根本性障碍，在人体姿态领域仍未被使用：缺乏预训练的流先验，以及关节姿态的非欧几里得本质。我们通过提出 PoseD-Flow 同时克服了这两个障碍，该框架将黎曼流匹配（RFM）与免训练引导统一起来，用于三维人体姿态恢复。PoseD-Flow 包含两项贡献：(i) PoseRFM，首个直接定义在关节旋转乘积流形上的人体姿态 RFM 模型；(ii) 黎曼 D-Flow，一种有原理依据的引导机制，通过对 ODE 采样动力学求导，在推理阶段对 PoseRFM 施加条件，而无需任何任务特定的训练。我们的理论分析表明，所诱导的动力学由数据协方差与流形曲率共同塑造，从而产生对真实姿态的偏好。在姿态补全、去噪与逆运动学任务上，PoseD-Flow 均取得了新的最优性能，尤其是在噪声、遮挡与部分观测条件下。
英文原文摘要 1133 字符
Generative pose priors have recently emerged as a powerful tool for inference under occlusion or noise. Yet today's strongest generative paradigm, flow matching, remains unused for human pose due to two fundamental barriers: the absence of a pre-trained flow prior and the non-Euclidean nature of articulated poses. We overcome both by introducing PoseD-Flow, a novel framework to unify Riemannian Flow Matching (RFM) with training-free guidance for 3D human pose recovery. PoseD-Flow is composed of two contributions: (i) PoseRFM, the first RFM model of human pose, defined directly on the product manifold of joint rotations, and (ii) Riemannian D-Flow, a principled guidance mechanism that, by differentiating through its ODE sampling dynamics, conditions PoseRFM at inference without any task-specific training. Our theoretical analysis shows that the induced dynamics are shaped by data covariance and manifold curvature, yielding a bias toward realistic poses. Across pose completion, denoising, and inverse kinematics, PoseD-Flow establishes new state of the art, particularly under noise, occlusion, and partial observations.