R-4R

The R4R (room-for-room) dataset is used to evaluate vision-and-language navigation tasks by providing a benchmark for agents to navigate between rooms based on natural language instructions.

Papers using R-4R (3)

PanoGen++: Domain-Adapted Text-Guided Panoramic Environment Generation for Vision-and-Language Navigation2025NavHint: Vision and Language Navigation Agent with a Hint Generator2024DELAN: Dual-Level Alignment for Vision-and-Language Navigation by Cross-Modal Contrastive Learning2024