CVDN

The CVDN dataset/benchmark is used to evaluate vision-and-language navigation tasks that involve navigating real-world scenes based on dialogue instructions.

Papers using CVDN (8)

PanoGen++: Domain-Adapted Text-Guided Panoramic Environment Generation for Vision-and-Language Navigation2025 · 9 citesImplicit Geometry Representations for Vision-and-Language Navigation from Web Videos2026History Aware Multimodal Transformer for Vision-and-Language Navigation2021 · 21 citesVision-Dialog Navigation by Exploring Cross-modal Memory2020 · 3 citesESceme: Vision-and-Language Navigation with Episodic Scene Memory2023 · 1 citesGrounded Entity-Landmark Adaptive Pre-training for Vision-and-Language Navigation2023DELAN: Dual-Level Alignment for Vision-and-Language Navigation by Cross-Modal Contrastive Learning2024Why Only Text: Empowering Vision-and-Language Navigation with Multi-modal Prompts2024