论文arXiv 2611.251342 天前
GEMS框架融合生成子目标与经验记忆提升长程规划成功率
本文由 AI 根据公开内容整理,可能有误,请以原文为准:GEMS: Generation and Experience Memory for Subgoal-Guided Long-Horizon World Model Planning(arXiv 2611.25134)
核心内容
针对视觉世界模型在长程机器人任务中预测误差累积及远距离目标表征指导失效的问题,研究提出GEMS(Generation and Experience Memory for Subgoal-Guided Planning)框架。该方法通过生成时序结构化子目标并结合潜在经验记忆(LEM)来引导动作搜索与状态选择,引入距离门控机制动态评估预测结果。实验覆盖操作、导航及抓取等六类任务,验证了其在冻结世界模型下增强长程规划的有效性。
技术方案上,GEMS利用生成的子目标弥补纯记忆检索的覆盖局限,同时用真实经验状态修正生成模型的幻觉与位置偏差。量化评测显示,该框架在Cube、Push-T和Meta-World基准上的成功率分别为92.8%、68.4%和54.1%,相比当前最强世界模型规划基线分别提升24.0、7.3和22.2个百分点,显著解决了长程推理中的漂移问题,为具身智能复杂任务执行提供了新范式。
关键看点
- 核心架构:GEMS融合生成式时序子目标与潜在经验记忆,通过距离门控机制动态切换评估对象,解决长程规划中误差累积与目标指引失效问题。
- 性能突破:在Cube、Push-T及Meta-World任务中成功率分别达92.8%、68.4%和54.1%,较SOTA基线最高提升24.0个百分点,验证了混合引导策略的有效性。
- 技术优势:无需重训世界模型即可增强长程能力,利用真实经验状态修正生成子目标的幻觉与位置偏差,克服了纯生成或纯记忆方法的固有局限。
工程与产业观察
GEMS证明了在不微调世界模型的前提下,通过外挂“生成子目标+经验记忆”双路引导可有效突破长程规划瓶颈。这对具身智能工程落地具有重要参考价值:在基础模型迭代周期长、算力昂贵的背景下,利用结构化记忆与推理时搜索增强,是低成本提升机器人复杂任务泛化能力的务实路径。
英文摘要 (Abstract)
Visual world models enable effective short-horizon goal-conditioned planning, such as reaching nearby targets and pushing objects over short distances, by predicting candidate action outcomes and comparing their representations with a goal representation. However, extending this approach to long-horizon tasks is challenging: prediction errors accumulate over extended rollouts, while latent-space similarity to distant goals provides limited guidance for ranking candidate outcomes. Existing methods address these limitations by introducing intermediate subgoals, either generated by learned models or selected from experience memory. Generated subgoals can contain hallucinations or be inaccurate in aspects such as object positions and timing. Memory-based approaches are constrained by experience coverage and may lack precise guidance for selecting appropriate intermediate states from stored observations. We introduce GEMS, which generates temporally structured subgoals to guide action search and state selection from Latent Experience Memory (LEM). A distance gate determines whether predicted outcomes are evaluated against the generated subgoal or the selected memory state. Experiments across six tasks and task suites spanning manipulation, navigation, and reaching demonstrate the effectiveness of GEMS for long-horizon planning. GEMS outperforms state-of-the-art world-model planning baselines across different manipulation settings, achieving success rates of 92.8% on Cube, 68.4% on Push-T, and 54.1% on Meta-World. It improves over the strongest evaluated baseline on these tasks by 24.0, 7.3, and 22.2 percentage points, respectively. These results demonstrate that generated subgoals and experience memory can jointly strengthen long-horizon planning with a frozen world model.
作者:Yangchen Zhao, Ali Payani, Caiwen Ding
外部来源
本页事实对应的公开页面,以原文为准。(点击将离开本站)