论文/arXiv 2610.21515·13 天前

TRACE：用于人形运动扩散的类型化参考感知对话式编辑

TRACE用类型化可测运动学量连接语言请求与参考运动，在Unitree G1运动生成和多轮编辑评估中提升对齐与质量。

AI 整理

• 本体 SDK 与控制
• 评测与安全

宇树 G1

原文链接：TRACE: Typed Reference-Aware Conversational Editing for Humanoid Motion Diffusion（arXiv 2610.21515 · 发布日期：2026-10-15）

TRACE: Typed Reference-Aware Conversational Editing for Humanoid Motion Diffusion

TRACE用类型化可测运动学量连接语言请求与参考运动，在Unitree G1运动生成和多轮编辑评估中提升对齐与质量。

核心要点

• 动机是将欠明确的自然语言请求落到可修改的具体运动结构上。
• 方法从配对训练运动预测支撑、根部和手部事件，并与原文本共同条件化扩散生成器。
• 引入重力参考状态和轨迹—接触目标以提升运动学一致性，编辑时用参考条件扩散编辑器实现相对修改。
• 定量评估与人工评估显示语义对齐和运动质量优于对比方法，多轮编辑更可靠。

验证范围与局限：验证范围为Unitree G1的运动生成与参考感知多轮编辑；摘要未说明仿真或真机。

英文摘要 (Abstract)

Language-conditioned humanoid motion generation enables users to create and refine robot movements through natural language. The central challenge is to connect an underspecified request to concrete motion structure that can subsequently be revised relative to an accepted reference. We present TRACE, a generation and editing framework for the Unitree G1 humanoid robot built around typed, measurable kinematic quantities. Our insight is to ground generation conditions in paired training motions and editing goals in the current motion. A language model learns to predict support, root, and hand events from motion-derived supervision; their time-aligned encodings complement the original text in a diffusion generator. Gravity-referenced states and trajectory--contact objectives improve kinematic consistency. For editing, typed requests define body--time scopes and source-relative goals. A separate reference-conditioned diffusion editor uses these goals and revised events to realize changes, while sampling-time projection preserves protected state coordinates. Quantitative evaluation and human assessment show improved semantic alignment and motion quality over the compared methods. Controlled ablations support the value of motion-grounded conditioning and kinematic consistency, while multi-turn editing evaluations demonstrate more reliable execution of successive relative requests under shared preservation constraints. Together, these results support typed kinematic structure as a common basis for motion generation and reference-aware revision.

作者：Zejian Kang, Yifu Luo, Boyang Shen, Jingxuan Li, Zekai Chen, Zekun Shi, Yongxue Xu, Zhenyu Hou, Bingqian Wu, Shaolei Zhang

来源

本页事实对应的公开页面，以原文为准。

1. arXiv 2610.21515：TRACE: Typed Reference-Aware Conversational Editing for Humanoid Motion Diffusion访问 2026-10-21