VG-Steer:基于上下文对齐抽象的可视-几何策略引导
VG-Steer: Visual-Geometric Policy Steering from In-Context Aligned Abstractions
Arthur Bucker · Pramod Anantharam · Jonathan Francis · Jean Oh
中文摘要
在有限真实数据下将机器人策略适配到分布外(OOD)场景,仍是通用机器人面临的关键瓶颈。机器人数据的多模态性与稀缺性导致端到端方法容易过拟合,并在小幅分布偏移下失效。测试时自适应可在推理阶段修正冻结策略以缓解该问题,但多数方法将修正建立在对机器人数据集隐式学习的信号上,因而同样易受策略所遭遇的分布偏移影响。另一些利用显式空间信息的方法通常只能做粗粒度修正,缺少多数操作任务所需的细微且与任务相关的运动。VG-Steer 通过迁移成功且分布内(in-distribution)轨迹中的细微运动来修正现有策略,这些运动沿有意义的几何轴(即物体关系距离对)被捕获。VG-Steer 利用大型视觉模型(LVM)与视觉-语言模型(VLM)推理的语义理解,将任务相关实体对及其物体关系距离投影组合成单一的定制化视觉-几何引导信号。在仿真基准中,它能高效修正流匹配(flow-matching)策略在姿态与任务目标偏移下的表现;在真实机器人实验中,它扩大了冻结策略覆盖的工作空间区域,在训练位置外一个网格步长的线缆放置任务上成功率提升一倍以上,并能将单任务策略引导至从未训练过的物体,同时几乎消除物体混淆。项目网站:https://arthurfenderbucker.github.io/VG-Steer/
关键要点
- 01问题:机器人数据稀缺且多模态,端到端策略在分布偏移下易过拟合失效。
- 02方法:利用LVM与VLM推理,将物体关系距离对组合为视觉-几何引导信号修正冻结策略。
- 03结果:仿真中高效修正流匹配策略在姿态与任务目标偏移下的表现。
- 04结果:真实机器人上线缆放置成功率提升一倍以上,并几乎消除物体混淆。
- 05局限:修正依赖成功分布内轨迹,未提及对无此类轨迹场景的适用性。
解读
尚无解读。
原始英文摘要
arXiv:2611.25199v1 Announce Type: new Abstract: Adapting robot policies to out-of-distribution (OOD) settings with limited real-world data remains a key bottleneck to generalist robots. The multimodality and scarcity of robotics data lead end-to-end approaches to overfit and break under small distribution shifts. Test-time adaptation can alleviate this by correcting frozen policies at inference, but most methods ground their correction in a signal learned implicitly from robotic datasets, making it susceptible to the same shifts as the policy it corrects. Recent methods that instead exploit explicit spatial information are typically limited to coarse corrections, missing the nuanced, task-relevant motion most manipulation tasks require. We propose VG-Steer, which corrects existing policies by transferring the nuanced motion of successful in-distribution rollouts, captured along meaningful geometric axes, known here as object-relational distance pairs. VG-Steer leverages the semantic understanding of Large Vision Models (LVMs) and Vision-Language Models (VLMs) reasoning, which combines the task-relevant entity pairs and their object-relational distance projections into a single tailored visual-geometric guidance signal. In a simulation benchmark, it efficiently corrects flow-matching policies under pose and task-goal shifts; on a real robot experiment, it widens the region of the workspace the frozen policy covers, more than doubling its success rate on cable placements one grid step outside the training positions, and steers a single-task policy toward objects it was never trained on while nearly eliminating object confusion. Project website: https://arthurfenderbucker.github.io/VG-Steer/