Research map / Vision-language and multimodal / Robot action and navigation
Scaling Data Generation in Vision-and-Language Navigation
Zun Wang, Jialu Li, Yicong Hong, Yi Wang, Qi Wu, Mohit Bansal, Stephen Gould, Hao Tan, Yu Qiao
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 |
|---|---|---|---|
| use dense, fully traversable navigation graphs instead of AutoVLN graphs (ablation) | DUET on R2R · R2R val-unseen | SR: 72.75 -> 76.01 | Improved |
| increase number of augmented scenes from 200 to 800 at 700k samples (ablation) | DUET on R2R · R2R val-unseen | SR: 74.84 -> 76.25 | Improved |
| increase samples per scene from 700k to 2890k with 800 scenes (ablation) | DUET on R2R · R2R val-unseen | SR: 76.25 -> 76.59 | Improved |
| recover faulty rendered images in augmented data (ablation) | Gibson augmented data · R2R val-unseen | SR: 76.33 -> 78.08 | Improved |
| add Gibson scenes with faulty images to HM3D data (ablation) | DUET on R2R · R2R val-unseen | SR: 76.59 -> 76.33 | No clear effect |
| use EnvDrop LSTM speaker instead of fine-tuned GPT-2 for instructions (ablation) | DUET on R2R · R2R val-unseen | SR: 75.82 -> 76.59 | Improved |
| keep augmented HM-E data in fine-tuning (ablation) | DUET on R2R · R2R val-unseen | SR: 76.59 -> 79.10 | Improved |
| include PREV data in fine-tuning (ablation) | DUET on R2R · R2R val-unseen | SR: 79.10 -> 77.10 | Hurt |
| add MLM and SAP pre-training tasks (ablation) | DUET on R2R · R2R val-unseen | SR: 62.62 -> 76.59 | Improved |
| add MRM to MLM and SAP pre-training (ablation) | DUET on R2R · R2R val-unseen | SR: 76.59 -> 75.82 | Hurt |
| pre-train with ScaleVLN data in addition to R2R and PREV (no fine-tune) (ablation) | DUET · R2R val-unseen | SR: 47.42 -> 57.00 | Improved |
| scale up with ScaleVLN data on REVERIE | DUET · REVERIE test-unseen | SR: 52.51 -> 56.13 | 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.
- DUET — DUET: Cross-Modal Semantic Grounding for Contrastive Zero-Shot Learning (AAAI 2023) · better on 3 of 3 shared results
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
- NavGPT: Explicit Reasoning in Vision-and-Language Navigation with Large Language Models (AAAI 2024 · similarity 0.91)
- Scene Map-based Prompt Tuning for Navigation Instruction Generation (CVPR 2025 · similarity 0.89)
- Habitat Synthetic Scenes Dataset (HSSD-200): An Analysis of 3D Scene Scale and Realism Tradeoffs for ObjectGoal Navigation (CVPR 2024 · similarity 0.89)
- General Scene Adaptation for Vision-and-Language Navigation (ICLR 2025 · similarity 0.89)
- Hierarchical Semantic-Augmented Navigation: Optimal Transport and Graph-Driven Reasoning for Vision-Language Navigation (NeurIPS 2025 · similarity 0.89)
- STRIDER: Navigation via Instruction-Aligned Structural Decision Space Optimization (NeurIPS 2025 · 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.
