R-2R

The R2R dataset is a benchmark for Vision-Language Navigation that contains trajectory instruction data from indoor environments, used to evaluate the performance of navigation models.

Papers using R-2R (3)

pFedNavi: Structure-Aware Personalized Federated Vision-Language Navigation for Embodied AI2026FedVLN: Privacy-preserving Federated Vision-and-Language Navigation2022Navigation as Attackers Wish? Towards Building Robust Embodied Agents under Federated Learning2022