发表机构
University of Fukui（福井大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对超大规模室内环境中资源受限的多目标导航，提出战略Transformer，将任务建模为集合定向问题，结合几何注意力与2-opt局部优化，实现94倍加速，在ProcTHOR上优于基线。
AI 中文摘要
在广阔室内环境（面积超过2,000平方米）中的资源受限多目标导航，对效率和战略规划提出了重大挑战。为解决这一问题，我们将该任务重新表述为集合定向问题（SOP），在资源约束下提供一个优化框架，其中开发由SOP模型控制，而探索由独立的启发式切换器管理。传统基线方法要么受限于僵化的规划，要么表现出短视行为。为克服这些局限并解决SOP在实时导航中的NP-hard计算挑战，我们开发了战略Transformer。这种轻量级架构充当优先级规划器，将专家组合逻辑内化到可预测的41.03毫秒前向传播中，同时减少师生信息不对称。引入几何注意力偏置使网络能够有效建模长程结构依赖。通过将Transformer的宏观规划与有界候选图上的有界迭代2-opt局部细化相结合，我们的框架相比重型元启发式方法实现了94倍加速，确保适合机载部署的有界延迟推理。在ProcTHOR上的实验验证了我们的方法成功弥合了探索与开发之间的差距，在按路径长度加权的进度（PPL）指标下优于精心重新实现的基线，并为可扩展、资源受限的导航建立了新基准。
英文摘要
Resource-constrained multi-object navigation in vast indoor environments ($>2,000\text{ m}^2$) poses significant challenges for efficiency and strategic planning. To tackle this, we reformulate the task as a Set Orienteering Problem (SOP), providing an optimization framework under resource constraints where exploitation is governed by the SOP model and exploration is managed by a separate heuristic switcher. Conventional baselines suffer from either rigid planning or myopic behaviors. To overcome these limitations and resolve the NP-hard computational challenges of SOP for real-time navigation, we develop the Strategic Transformer. This lightweight architecture functions as a priority planner that internalizes expert combinatorial logic into a predictable $41.03\text{ ms}$ forward pass while reducing teacher-student information asymmetry. Incorporating geometric attention biases allows the network to effectively model long-range structural dependencies. By coupling the Transformer's macro-plan with a bounded iterative 2-opt local refinement on a capped candidate graph, our framework achieves a $94\times$ speedup compared to heavy meta-heuristics, ensuring bounded-latency inference suitable for onboard deployment. Experiments on ProcTHOR validate that our method successfully bridges the gap between exploration and exploitation, outperforming carefully re-implemented baselines under Progress weighted by Path Length (PPL) and establishing a new benchmark for scalable, resource-constrained navigation.
Comments8 pages, 6 figures, technical report
