Research map / Vision-language and multimodal / Robot action and navigation
Dual-Anchoring: Addressing State Drift in Vision-Language Navigation
Kangyi Wu, Pengna Li, Kailin Lyu, Xi Lin, Lin Zhao, Qingrong He, Jinjun Wang, Jianyi Liu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add Instruction Progress Anchoring (progress text co-training) (ablation) | StreamVLN · R2R-CE val unseen (R2R+RxR+EnvDrop) | SR (%): 40.8 -> 45.4 | Improved |
| add Memory Landmark Anchoring (landmark world model) (ablation) | StreamVLN · R2R-CE val unseen (R2R+RxR+EnvDrop) | NE: 6.49 -> 6.01 | Improved |
| add Memory Landmark Anchoring (landmark world model) (ablation) | StreamVLN · R2R-CE val unseen (R2R+RxR+EnvDrop) | SPL (%): 37.4 -> 43.3 | Improved |
| add both IPA and MLA (ablation) | StreamVLN · R2R-CE val unseen (R2R+RxR+EnvDrop) | SR (%): 40.8 -> 49.5 | Improved |
| add Instruction Progress Anchoring (ablation) | StreamVLN · R2R-CE val unseen (all datasets) | SR (%): 56.9 -> 60.8 | Improved |
| add Memory Landmark Anchoring (ablation) | StreamVLN · R2R-CE val unseen (all datasets) | SR (%): 56.9 -> 62.5 | Improved |
| add both IPA and MLA (ablation) | StreamVLN · R2R-CE val unseen (all datasets) | SPL (%): 51.9 -> 62.1 | Improved |
| add Dual-Anchoring framework | StreamVLN · RxR-CE val unseen | SR (%): 52.9 -> 61.7 | Improved |
| add Dual-Anchoring framework on long trajectories | StreamVLN · R2R-CE val unseen by trajectory length | SPL relative improvement: +12.6% -> +33.2% | Improved |
| use Visual Kinematics Prompting (frame index and action overlay) for progress annotation (ablation) | Qwen3-VL annotator · Progress description data | Hallucination rate (%): 8.13 -> 6.04 | Improved |
| use Visual Kinematics Prompting for progress annotation (ablation) | Qwen3-VL annotator · Progress description data | Logical consistency score: 1.71 -> 4.26 | Improved |
| mine landmark frames with ordered temporal grounding instead of random sampling (ablation) | Qwen3-VL · Grounded landmark data | Landmark presence rate (%): 13.9 -> 75.6 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- ETPNav — ETPNav: Evolving Topological Planning for Vision-Language Navigation in Continuous Environments (TPAMI 2024) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Instruction Decomposition and Action Alignment for Vision-Language Navigation (ICML 2026 · similarity 0.92)
- Progress-Think: Semantic Progress Reasoning for Vision-Language Navigation (CVPR 2026 · similarity 0.92)
- Towards Long-Horizon Vision-Language Navigation: Platform, Benchmark and Method (CVPR 2025 · similarity 0.90)
- STRIDER: Navigation via Instruction-Aligned Structural Decision Space Optimization (NeurIPS 2025 · similarity 0.89)
- Scene Map-based Prompt Tuning for Navigation Instruction Generation (CVPR 2025 · similarity 0.89)
- NavGPT: Explicit Reasoning in Vision-and-Language Navigation with Large Language Models (AAAI 2024 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
