用3D世界模型+蒙特卡洛树搜索，让机器人记住被遮挡的物体位置
3D-Anchored Lookahead Planning for Persistent Robotic Scene Memory via World-Model-Based MCTS
- 以3D世界锚点保持空间记忆，支持遮挡后重规划
- 在需记忆的步骤中成功率提升至65%，比基线高64.5个百分点
- 解决连续控制中蒙特卡洛树搜索的四个结构性问题
我们提出3D-Anchored Lookahead Planning（3D-ALP），一种基于世界模型的蒙特卡洛树搜索（MCTS）系统，用于机器人操作中的持续场景记忆。与仅依赖当前相机帧的反应式策略不同，3D-ALP维护一个跨遮挡的相机到世界（c2w）锚点，实现对不可见物体位置的准确重规划。在需要空间记忆的五步序列抓取任务（实验E3）中，3D-ALP在需记忆步骤的成功率为0.650 ± 0.109，远高于贪心反应基线的0.006 ± 0.008（Δ=+0.645），第5步成功率达0.822，而基线为0.000。消融实验（30个轨迹，3个随机种子）表明，树搜索带来的空间记忆是主要贡献（+0.533，占总增益82%），更深层前瞻也有额外收益（+0.111，17%）。此外，我们识别并解决了将UCT-MCTS应用于连续机器人操作时的四个结构性失败模式。
原文摘要 · Abstract (English)
We present 3D-Anchored Lookahead Planning (3D-ALP), a System 2 reasoning engine for robotic manipulation that combines Monte Carlo Tree Search (MCTS) with a 3D-consistent world model as the rollout oracle. Unlike reactive policies that evaluate actions from the current camera frame only, 3D-ALP maintains a persistent camera-to-world (c2w) anchor that survives occlusion, enabling accurate replanning to object positions that are no longer directly observable. On a 5-step sequential reach task requiring spatial memory (Experiment E3), 3D-ALP achieves 0.650 0.109 success rate on memory-required steps versus 0.006 0.008 for a greedy reactive baseline (Δ=+0.645), while step 5 success reaches 0.822 against 0.000 for greedy. An ablation study (30 episodes, 3 seeds) isolates tree search spatial memory as the primary driver (+0.533, 82% of gain) with additional benefit from deeper lookahead (+0.111, 17%). We also identify and resolve four structural failure modes in applying UCT-MCTS (Upper Confidence Bounds applied to Trees [10]) to continuous robotic manipulation.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。