arXiv cs.MA· I-tak Ieong, Hao Tang·· 12 小时前AI 评分29
HGWM：用分层图引导世界模型与图匹配实现零样本物体导航
HGWM: Hierarchical Graph-Guided World Model for Zero-Shot Object Navigation via Graph Matching
AI 导读
HGWM 将 LLM 生成的房间锚点先验与持久空间语义记忆结合，通过粗粒度视觉语义打分与细粒度符号匹配两阶段，在局部动作选择前融合 LLM 方向建议。在 HM3D v0.1 上取得 62.6% SR 和 31.5% SPL，高于 WMNav 的 58.1% 和 31.2%；MP3D 上为 46.8% 和 17.9%。
正文 · 原文
Abstract:Zero-shot object navigation requires an embodied agent to choose where to search while retaining evidence from previous observations. HGWM couples an LLM-generated room-anchor prior with persistent spatial-semantic memory. A coarse stage assigns visual-semantic scores to candidate headings on a 0-10 scale; a fine stage measures symbolic compatibility between observed objects, room context, and the goal prior. Phase-dependent fusion combines these scores with an LLM direction recommendation before local action selection. On HM3D v0.1, the reported HGWM configuration achieves 62.6% success rate (SR) and 31.5% success weighted by path length (SPL), compared with 58.1% and 31.2% for WMNav. On MP3D, the corresponding values are 46.8% and 17.9%, compared with 45.4% and 17.2%. The results indicate higher success point estimates and modest path-efficiency gains. On HM3D v0.2, the full model obtains 74.4% SR and 31.9% SPL. Additional component ablations reduce SR by 0.9-2.3 percentage points and SPL by 0.6-1.5 points. These are author-reported aggregate measurements. The accompanying implementation repairs pass offline tests; episode-level rerun records are not available. Evaluation is confined to simulation.
| Subjects: | Robotics (cs.RO); Multiagent Systems (cs.MA) |
| Cite as: | arXiv:2610.32554 [cs.RO] |
| (or arXiv:2610.32554v1 [cs.RO] for this version) | |
| https://doi.org/10.48550/arXiv.2610.32554 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: I Tak Ieong [view email]
[v1]
Fri, 23 Oct 2026 12:35:26 UTC (4,399 KB)
来源：arXiv cs.MA · arxiv.org