Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2604.16993 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908976065544192 |
|---|---|
| author | Wen, Jiawen Sun, Penglei Zhang, Wenjie Qiu, Suixuan Xu, Weisheng Yang, Xiaofei Chu, Xiaowen |
| author_facet | Wen, Jiawen Sun, Penglei Zhang, Wenjie Qiu, Suixuan Xu, Weisheng Yang, Xiaofei Chu, Xiaowen |
| contents | As embodied AI transitions to real-world deployment, the success of the Vision-and-Language Navigation (VLN) task tends to evolve from mere reachability to social compliance. However, current agents suffer from a "goal-driven trap", prioritizing physical geometry ("can I go?") over semantic rules ("may I go?"), frequently overlooking subtle regulatory constraints. To bridge this gap, we establish Rule-VLN, the first large-scale urban benchmark for rule-compliant navigation. Spanning a massive 29k-node environment, it injects 177 diverse regulatory categories into 8k constrained nodes across four curriculum levels, challenging agents with fine-grained visual and behavioral constraints. We further propose the Semantic Navigation Rectification Module (SNRM), a universal, zero-shot module designed to equip pre-trained agents with safety awareness. SNRM integrates a coarse-to-fine visual perception VLM framework with an epistemic mental map for dynamic detour planning. Experiments demonstrate that while Rule-VLN challenges state-of-the-art models, SNRM significantly restores navigation capabilities, reducing CVR by 19.26% and boosting TC by 5.97%. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_16993 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Rule-VLN: Bridging Perception and Compliance via Semantic Reasoning and Geometric Rectification Wen, Jiawen Sun, Penglei Zhang, Wenjie Qiu, Suixuan Xu, Weisheng Yang, Xiaofei Chu, Xiaowen Artificial Intelligence Computer Vision and Pattern Recognition Robotics As embodied AI transitions to real-world deployment, the success of the Vision-and-Language Navigation (VLN) task tends to evolve from mere reachability to social compliance. However, current agents suffer from a "goal-driven trap", prioritizing physical geometry ("can I go?") over semantic rules ("may I go?"), frequently overlooking subtle regulatory constraints. To bridge this gap, we establish Rule-VLN, the first large-scale urban benchmark for rule-compliant navigation. Spanning a massive 29k-node environment, it injects 177 diverse regulatory categories into 8k constrained nodes across four curriculum levels, challenging agents with fine-grained visual and behavioral constraints. We further propose the Semantic Navigation Rectification Module (SNRM), a universal, zero-shot module designed to equip pre-trained agents with safety awareness. SNRM integrates a coarse-to-fine visual perception VLM framework with an epistemic mental map for dynamic detour planning. Experiments demonstrate that while Rule-VLN challenges state-of-the-art models, SNRM significantly restores navigation capabilities, reducing CVR by 19.26% and boosting TC by 5.97%. |
| title | Rule-VLN: Bridging Perception and Compliance via Semantic Reasoning and Geometric Rectification |
| topic | Artificial Intelligence Computer Vision and Pattern Recognition Robotics |
| url | https://arxiv.org/abs/2604.16993 |