R-2R

The R2R dataset/benchmark contains a collection of navigation tasks that evaluate Vision-Language Navigation (VLN) systems by requiring agents to follow natural language instructions to navigate through real-world environments.

Papers using R-2R (11)

Does Peer Observation Help? Vision-sharing Collaboration For Vision-language Navigation2026Fine-Grained Instruction-Guided Graph Reasoning for Vision-and-Language Navigation2025Think Hierarchically, Act Dynamically: Hierarchical Multi-modal Fusion and Reasoning for Vision-and-Language Navigation2025CoNav: Collaborative Cross-Modal Reasoning for Embodied Navigation2025Landmark-Guided Knowledge for Vision-and-Language Navigation2025Global Commander and Local Operative: A Dual-Agent Framework for Scene Navigation2026Implicit Geometry Representations for Vision-and-Language Navigation from Web Videos2026Trajectory-Diversity-Driven Robust Vision-and-Language Navigation2026Does Peer Observation Help? Vision-Sharing Collaboration for Vision-Language Navigation2026Structured Observation Language for Efficient and Generalizable Vision-Language Navigation2026Vision-and-Language Navigation with Analogical Textual Descriptions in LLMs2025