Fetching the paper…

DELAN: Dual-Level Alignment for Vision-and-Language Navigation by Cross-Modal Contrastive Learning · Around