外观
论文arXiv 2610.1454118 天前
GG-Nav：面向目标的 Grounding Chain-of-Thought 激发导航基础模型推理
针对目标导向导航表现不足，GG-Nav以监督微调学习结构化CoT，并在多个基准和真机上验证。
宇树 Go2原文链接：GG-Nav: Goal-oriented Grounding Chain-of-Thought Elicits Reasoning in Navigation Foundation Models（arXiv 2610.14541 · 发布日期：2026-10-10）
GG-Nav: Goal-oriented Grounding Chain-of-Thought Elicits Reasoning in Navigation Foundation Models
针对目标导向导航表现不足，GG-Nav以监督微调学习结构化CoT，并在多个基准和真机上验证。
核心要点
- 动机研究指出，现有导航基础模型对流式观测利用不足，且目标定位不够精确。
- GG-Nav通过结构化CoT显式定位目标相关实例、累积grounding证据并支持决策。
- 方法包含Observation-to-Evidence与Evidence-to-Decision两阶段，利用中间帧证据生成动作或澄清问题。
- 实验显示其在HM3D ObjectNav、VL-LN和CoIN成功率优于现有方法，并在Unitree Go2上展示零样本泛化。
验证范围与局限：验证范围包括HM3D ObjectNav、VL-LN、CoIN基准，以及一台Unitree Go2机器人在未见真实环境中的部署。
英文摘要 (Abstract)
Emerging strong Vision-Language Models (VLMs) have advanced navigation foundation models. However, these models still have unsatisfactory performance on goal-oriented navigation settings. In this paper, we first conduct a motivating study revealing that the key issue lies in the insufficient utilization of streaming observations and imprecise target localization. To tackle these challenges, we propose GG-Nav, a simple yet effective navigation framework that learns a structured chain-of-thought (CoT) mechanism via supervised fine-tuning to explicitly localize goal-relevant instances, accumulate grounding evidence, and support more reliable decision-making. Specifically, the structured CoT is designed with two key stages: (i) Observation-to-Evidence, which reasons over all intermediate frames between adjacent decision steps to retain useful goal-related evidence; and (ii) Evidence-to-Decision, which integrates the current observation with accumulated evidence to ground the target and support the generation of navigation actions or clarification questions. Extensive experiments show that GG-Nav outperforms the state-of-the-art methods in terms of success rates on the HM3D ObjectNav, VL-LN, and CoIN benchmarks by 3.1%, 16.6%, and 5.3%, respectively. Ablation studies further show that accumulating evidence from intermediate observations reduces evidence-missing failures by 11.6%, while explicit grounding reduces stopping errors by 49.4%. We also deploy GG-Nav on a Unitree Go2 robot, demonstrating its zero-shot generalization capability in unseen real-world environments.
作者：Shaohao Zhu, Wensi Huang, Siqi Zhang, Yuqiang Yang, Meng Wei, Chenyang Wan, Tianle Liu, Tai Wang, Jinming Xu
来源
本页事实对应的公开页面，以原文为准。