While large vision-language models (VLMs) show promise for object goal navigation, current methods still struggle with low success rates and inefficient localization of unseen objects--failures primarily attributed to weak temporal-spatial reasoning. Meanwhile, recent attempts to inject reasoning into VLM-based agents improve success rates but incur substantial computational overhead. To address both the ineffectiveness and inefficiency of existing approaches, we introduce Hydra-Nav, a unified VLM architecture that adaptively switches between a deliberative slow system for analyzing exploration history and formulating high-level plans, and a reactive fast system for efficient execution. We train Hydra-Nav through a three-stage curriculum: (i) spatial-action alignment to strengthen trajectory planning, (ii) memory-reasoning integration to enhance temporal-spatial reasoning over long-horizon exploration, and (iii) iterative rejection fine-tuning to enable selective reasoning at critical decision points. Extensive experiments demonstrate that Hydra-Nav achieves state-of-the-art performance on the HM3D, MP3D, and OVON benchmarks, outperforming the second-best methods by 11.1%, 17.4%, and 21.2%, respectively. Furthermore, we introduce SOT (Success weighted by Operation Time), a new metric to measure search efficiency across VLMs with varying reasoning intensity. Results show that adaptive reasoning significantly enhances search efficiency over fixed-frequency baselines.

翻译：尽管大型视觉语言模型在目标导航任务中展现出潜力，但现有方法仍受限于较低的成功率及对未见目标物体的低效定位——这些失败主要归因于薄弱的时空推理能力。同时，近期为基于VLM的智能体注入推理能力的尝试虽提升了成功率，却带来了显著的计算开销。为同时解决现有方法的低效性与高耗能问题，我们提出了Hydra-Nav：一种统一的可自适应切换推理模式的VLM架构。该架构包含一个用于分析探索历史并制定高层规划的审慎慢速系统，以及一个用于高效执行的反应式快速系统。我们通过三阶段课程训练Hydra-Nav：（1）空间-动作对齐以强化轨迹规划能力；（2）记忆-推理融合以增强长时程探索中的时空推理能力；（3）迭代拒绝微调以实现关键决策点的选择性推理。大量实验表明，Hydra-Nav在HM3D、MP3D和OVON基准测试中均达到了最先进的性能，分别优于次优方法11.1%、17.4%和21.2%。此外，我们提出了SOT（操作时间加权成功率）这一新指标，用以衡量不同推理强度VLM的搜索效率。实验结果表明，自适应推理相比固定频率的基线方法能显著提升搜索效率。