MAG-Nav: Language-Driven Object Navigation Leveraging Memory-Reserved Active Grounding
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918116495196160 |
|---|---|
| author | Zhang, Weifan Li, Tingguang Liu, Yuzhen |
| author_facet | Zhang, Weifan Li, Tingguang Liu, Yuzhen |
| contents | Visual navigation in unknown environments based solely on natural language descriptions is a key capability for intelligent robots. In this work, we propose a navigation framework built upon off-the-shelf Visual Language Models (VLMs), enhanced with two human-inspired mechanisms: perspective-based active grounding, which dynamically adjusts the robot's viewpoint for improved visual inspection, and historical memory backtracking, which enables the system to retain and re-evaluate uncertain observations over time. Unlike existing approaches that passively rely on incidental visual inputs, our method actively optimizes perception and leverages memory to resolve ambiguity, significantly improving vision-language grounding in complex, unseen environments. Our framework operates in a zero-shot manner, achieving strong generalization to diverse and open-ended language descriptions without requiring labeled data or model fine-tuning. Experimental results on Habitat-Matterport 3D (HM3D) show that our method outperforms state-of-the-art approaches in language-driven object navigation. We further demonstrate its practicality through real-world deployment on a quadruped robot, achieving robust and effective navigation performance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_05021 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | MAG-Nav: Language-Driven Object Navigation Leveraging Memory-Reserved Active Grounding Zhang, Weifan Li, Tingguang Liu, Yuzhen Robotics Visual navigation in unknown environments based solely on natural language descriptions is a key capability for intelligent robots. In this work, we propose a navigation framework built upon off-the-shelf Visual Language Models (VLMs), enhanced with two human-inspired mechanisms: perspective-based active grounding, which dynamically adjusts the robot's viewpoint for improved visual inspection, and historical memory backtracking, which enables the system to retain and re-evaluate uncertain observations over time. Unlike existing approaches that passively rely on incidental visual inputs, our method actively optimizes perception and leverages memory to resolve ambiguity, significantly improving vision-language grounding in complex, unseen environments. Our framework operates in a zero-shot manner, achieving strong generalization to diverse and open-ended language descriptions without requiring labeled data or model fine-tuning. Experimental results on Habitat-Matterport 3D (HM3D) show that our method outperforms state-of-the-art approaches in language-driven object navigation. We further demonstrate its practicality through real-world deployment on a quadruped robot, achieving robust and effective navigation performance. |
| title | MAG-Nav: Language-Driven Object Navigation Leveraging Memory-Reserved Active Grounding |
| topic | Robotics |
| url | https://arxiv.org/abs/2508.05021 |