研究人员开发了 MoCapAnything V2，这是一个用于从单目视频进行 3D 动作捕捉的端到端系统，可适应任意骨架。该新框架将姿势和旋转预测整合到一个单一的、联合优化的可学习过程中，解决了先前分解管道的局限性。通过引入参考姿势-旋转对，该系统解决了旋转预测中的歧义，并在未见过的骨架上实现了显著提高的准确性，将旋转误差降低到约 6.54 度。该方法还通过直接从视频预测关节位置而无需中间网格来提高效率，从而使推理速度提高高达 20 倍。 AI
影响 通过实现对各种数字资产更准确、更高效的动作捕捉，增强了 3D 动画功能。
排序理由 该集群描述了一篇详细介绍了改进的 3D 动作捕捉方法的学术论文。
AI 生成摘要 · Google Gemini · 来自 3 个来源。 我们如何撰写摘要 →
arXiv:2604.28130v1 Announce Type: new Abstract: Recent methods for arbitrary-skeleton motion capture from monocular video follow a factorized pipeline, where a Video-to-Pose network predicts joint positions and an analytical inverse-kinematics (IK) stage recovers joint rotations.…
arXiv:2512.10881v2 Announce Type: replace Abstract: Motion capture now underpins content creation far beyond digital humans, yet most existing pipelines remain species- or template-specific. We formalize this gap as Category-Agnostic Motion Capture (CAMoCap): given a monocular vi…
Recent methods for arbitrary-skeleton motion capture from monocular video follow a factorized pipeline, where a Video-to-Pose network predicts joint positions and an analytical inverse-kinematics (IK) stage recovers joint rotations. While effective, this design is inherently limi…