Humanoid motion tracking is central to teleoperation and whole-body imitation, yet evaluation often disagrees with what people perceive in videos. Kinematic errors average per-frame pose differences but miss the physical artifacts that matter most, particularly unstable support and incorrect contacts such as foot skating and mistimed touch-downs. Meanwhile, widely used test suites are small and la

---
**📖 中文解读**
以上内容由AI翻译自英文原文，可能存在不准确之处。建议阅读[原文](https://arxiv.org/abs/2608.13555v1)获取最准确的信息。

---
🔗 **原文链接**: [HumanTracker: Towards Comprehensive and Human-Aligned Motion](https://arxiv.org/abs/2608.13555v1)
🏷️ **转载来源**: ArXiv cs.AI
> 本文由小九AI技术站翻译整理，内容版权归原作者所有。
👤 作者: Dairu Liu, Zekun Qi, Jiayu Zeng, Ruixi Yu, Yu Guan, Yintianrun Zhang, Xuchuan Chen, Sikai Liang, Zekai Li, Chenghuai Lin, Xinqiang Yu, Wenyao Zhang, He Wang, Li Yi

---
🐾 **小九锐评**

这篇论文来自arXiv预印本，虽然还没有经过同行评审，但选题方向值得关注。
建议先读中文摘要判断是否相关，再看全文细节。
Benchmark看多了容易麻木——跑分好不一定产品好用。这篇文章好在对分差有分析，不只是贴数据。

你对这个话题有什么看法？欢迎在评论区讨论 💬

> _转载自 ArXiv cs.AI，内容版权归原作者所有_

---
⏱️ 2026-08-14 14:02