Vision-and-Language Navigation on R2R (val unseen)
84Success Rate (SR)SRDF
Evaluation Results
| Method | Links | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| SRDFlarge-scale data augmentation=true2025.08 | 84 | — | 1.83 | 78 | — | — | — | 89 | — | — | — | — | |
| SkillNav (SRDF-Aug)large-scale data augmentation=true, visual encoder=InternViT-6B, action router=GPT-4o2025.08 | 84 | — | 1.79 | 78 | — | — | — | 89 | — | — | — | — | |
| SkillNav (ScaleVLN-Aug)large-scale data augmentation=true, visual encoder=CLIP-B/16, language backbone=BERT-base-uncased, action router=GPT-4o2025.08 | 83 | — | 1.97 | 77 | — | — | — | 89 | — | — | — | — | |
| DUET + ScaleVLN + Vision-SharingMethod Paradigm=Supervised Learning, Image Features=CLIP ViT-B16, Vision-Sharing=True, Spatial Overlap Detection=Oracle viewpoint IDs2026.03 | 82.46 | — | 2.12 | 73.39 | — | — | — | 89.1 | 12.31 | — | — | — | |
| DUET + ScaleVLN + Vision-SharingMethod Paradigm=Supervised Learning, Image Features=CLIP ViT-B16, Vision-Sharing=True2026.03 | 81.01 | — | 2.15 | 72.39 | — | — | — | 88.85 | 12.44 | — | — | — | |
| ScaleVLNSetting=Cross-modal backbone, Year=2023, multi-task pretraining=true2026.01 | 81 | — | 2.09 | 70 | — | — | — | 88 | 14.09 | — | — | — | |
| ScaleVLN2026.02 | 81 | — | 2.09 | 70 | — | — | — | 88 | 14.09 | — | — | — | |
| ScaleVLNSettings=Pretrain2026.02 | 81 | — | 2.09 | 70 | — | — | — | 88 | — | — | — | — | |
| DUET + ScaleVLMethod Paradigm=Supervised Learning, Image Features=CLIP ViT-B162026.03 | 79.4 | — | 2.38 | 69.97 | — | — | — | 87.48 | 12.4 | — | — | — | |
| Previous SOTA2024.09 | 79 | — | — | — | — | — | — | — | — | — | — | — | |
| GOAT-ATENASettings=Pretrain2026.02 | 79 | — | 2.27 | 69 | — | — | — | — | — | — | — | — | |
| ScaleVLNlarge-scale data augmentation=true2025.08 | 79 | — | 2.39 | 70 | — | — | — | 88 | — | — | — | — | |
| GOATSettings=Pretrain2026.02 | 78 | — | 2.4 | 68 | — | — | — | 85 | — | — | — | — | |
| GOAT2024.04 | 77.82 | — | 2.4 | 68.13 | — | — | — | 84.72 | — | — | — | — | |
| 3D Gaussian Map2026.05 | 77 | — | 2.43 | 66 | — | — | — | — | 14.83 | — | — | — | |
| DUET + Vision-SharingMethod Paradigm=Supervised Learning, Vision-Sharing=True, Spatial Overlap Detection=Oracle viewpoint IDs2026.03 | 76.23 | — | 2.63 | 62.9 | — | — | — | 85.06 | 14.62 | — | — | — | |
| P3Nav2026.03 | 76 | — | 2.79 | 65 | — | — | — | — | — | — | — | — | |
| SAMElarge-scale data augmentation=true2025.08 | 76 | — | 2.73 | 66 | — | — | — | — | — | — | — | — | |
| IDEABase Model=DUET2026.05 | 76 | — | 2.91 | 67 | — | — | — | — | 12.47 | — | — | — | |
| IDEABase Model=BEVBert2026.05 | 76 | — | 2.53 | 68 | — | — | — | — | 12.03 | — | — | — | |
| BEVBert + ATENABase Model=BEVBert, Feedback Setting=With Feedback2026.05 | 76 | — | 2.78 | 68 | — | — | — | — | 12.22 | — | — | — | |
| DUET + IDEABase Model=DUET, Feedback Setting=Without Feedback2026.05 | 76 | — | 2.91 | 67 | — | — | — | — | 12.47 | — | — | — | |
| BEVBert + IDEABase Model=BEVBert, Feedback Setting=Without Feedback2026.05 | 76 | — | 2.53 | 68 | — | — | — | — | 12.03 | — | — | — | |
| decay-onlyvariant=decay-only2026.07 | 76 | — | 2.59 | 65 | — | — | — | — | 13.29 | — | — | 742.61 | |
| decay+anti-loopvariant=decay+anti-loop2026.07 | 76 | — | 2.69 | 66 | — | — | — | — | 12.41 | — | — | 666.37 | |
| AirbertSpeaker=true, Follower=true2021.08 | 75.01 | 9.99 | 2.69 | 70 | — | — | — | 82.89 | — | — | — | — | |
| GridMM2024.04 | 75 | — | 2.83 | 64 | — | — | — | — | — | — | — | — | |
| BEVBert2024.04 | 75 | — | 2.81 | 64 | — | — | — | 84 | — | — | — | — | |
| GridMM2024.12 | 75 | — | — | 64 | — | — | — | — | — | — | — | — | |
| BEVBertSetting=Cross-modal backbone, Year=20222026.01 | 75 | — | 2.81 | 64 | — | — | — | 84 | 14.55 | — | — | — | |
| BEVBert2026.02 | 75 | — | 2.81 | 64 | — | — | — | 84 | 14.55 | — | — | — | |
| BEVBert2026.03 | 75 | — | 2.81 | 64 | — | — | — | — | — | — | — | — | |
| BEVBertMethod Paradigm=Supervised Learning2026.03 | 75 | — | 2.81 | 64 | — | — | — | 84 | 14.55 | — | — | — | |
| GridMM2026.03 | 75 | — | 2.83 | 64 | — | — | — | — | — | — | — | — | |
| BEVBert2026.03 | 75 | — | 2.81 | 64 | — | — | — | — | — | — | — | — | |
| BTK2026.03 | 75 | — | 2.69 | 64 | — | — | — | 84 | — | — | — | — | |
| BEVBERT2025.08 | 75 | — | 2.81 | 64 | — | — | — | 84 | — | — | — | — | |
| FSTTABase Model=DUET2026.05 | 75 | — | 3.03 | 62 | — | — | — | — | 14.64 | — | — | — | |
| BEVBertBase Model=BEVBert2026.05 | 75 | — | 2.81 | 64 | — | — | — | — | 14.55 | — | — | — | |
| SARBase Model=BEVBert2026.05 | 75 | — | 2.79 | 64 | — | — | — | — | 12.98 | — | — | — | |
| ViDABase Model=BEVBert2026.05 | 75 | — | 2.83 | 64 | — | — | — | — | 13.11 | — | — | — | |
| ReCAPBase Model=BEVBert2026.05 | 75 | — | 2.78 | 64 | — | — | — | — | 12.94 | — | — | — | |
| DUET + FeedTTABase Model=DUET, Feedback Setting=With Feedback2026.05 | 75 | — | 2.95 | 65 | — | — | — | — | 13.52 | — | — | — | |
| DUET + ATENABase Model=DUET, Feedback Setting=With Feedback2026.05 | 75 | — | 2.9 | 66 | — | — | — | — | 12.31 | — | — | — | |
| BEVBertBase Model=BEVBert, Feedback Setting=With Feedback2026.05 | 75 | — | 2.81 | 64 | — | — | — | — | 14.55 | — | — | — | |
| BEVBert + RLCFBase Model=BEVBert, Feedback Setting=With Feedback2026.05 | 75 | — | 2.76 | 64 | — | — | — | — | 14.11 | — | — | — | |
| BEVBert + FEEDTTABase Model=BEVBert, Feedback Setting=With Feedback2026.05 | 75 | — | 2.77 | 66 | — | — | — | — | 12.24 | — | — | — | |
| DUET + FSTTABase Model=DUET, Feedback Setting=Without Feedback2026.05 | 75 | — | 3.03 | 62 | — | — | — | — | 14.64 | — | — | — | |
| BEVBertBase Model=BEVBert, Feedback Setting=Without Feedback2026.05 | 75 | — | 2.81 | 64 | — | — | — | — | 14.55 | — | — | — | |
| BEVBert + ReCAPBase Model=BEVBert, Feedback Setting=Without Feedback2026.05 | 75 | — | 2.78 | 64 | — | — | — | — | 12.94 | — | — | — | |
| BEVBert2026.05 | 75 | — | 2.81 | 64 | — | — | — | — | 14.55 | — | — | — | |
| update-onlyvariant=update-only2026.07 | 75 | — | 2.86 | 63 | — | — | — | — | 13.92 | — | — | 891.12 | |
| full-modevariant=full-mode2026.07 | 75 | — | 2.8 | 64 | — | — | — | — | 13.77 | — | — | 1,000.84 | |
| DUET + Vision-SharingMethod Paradigm=Supervised Learning, Vision-Sharing=True2026.03 | 74.54 | — | 2.87 | 62.28 | — | — | — | 83.18 | 14.46 | — | — | — | |
| NavGPT-2Setting=Prompting-based LLM agents, Year=2024, additional synthetic trajectories=true2026.01 | 74 | — | 2.98 | 61 | — | — | — | 84 | 14.01 | — | — | — | |
| NavGPT22026.03 | 74 | — | 2.98 | 61 | — | — | — | 84 | — | — | — | — | |
| BSG2026.03 | 74 | — | 2.8 | 62 | — | — | — | — | — | — | — | — | |
| S-EGIU2026.04 | 74 | — | 3.37 | 61 | — | — | — | — | 13.33 | — | — | — | |
| TentBase Model=BEVBert2026.05 | 74 | — | 2.93 | 63 | — | — | — | — | 13.14 | — | — | — | |
| FSTTABase Model=BEVBert2026.05 | 74 | — | 2.89 | 63 | — | — | — | — | 13.96 | — | — | — | |
| BEVBert + FSTTABase Model=BEVBert, Feedback Setting=Without Feedback2026.05 | 74 | — | 2.89 | 63 | — | — | — | — | 13.96 | — | — | — | |
| BSG2026.05 | 74 | — | 2.89 | 62 | — | — | — | — | 14.9 | — | — | — | |
| BEVBert2026.07 | 74 | — | 2.83 | 64 | — | — | — | — | 14.52 | — | — | — | |
| GridMM (baseline)2026.07 | 74 | — | 2.83 | 64 | — | — | — | — | 13.27 | — | — | 937.99 | |
| VLN-BERTSpeaker=true, Follower=true2021.08 | 73.61 | 10 | 2.76 | 68 | — | — | — | 81.91 | — | — | — | — | |
| DSRG2024.04 | 73 | — | 3 | 62 | — | — | — | 81 | — | — | — | — | |
| CONSOLE2024.12 | 73 | — | 3 | 63 | — | — | — | — | — | — | — | — | |
| SUSA2024.12 | 73 | — | 3 | 64.8 | — | — | — | — | — | — | — | — | |
| KESU2026.03 | 73 | — | 2.96 | 62 | — | — | — | — | — | — | — | — | |
| CONSOLE2026.03 | 73 | — | 3 | 63 | — | — | — | — | — | — | — | — | |
| DUET + RLCFBase Model=DUET, Feedback Setting=With Feedback2026.05 | 73 | — | 3.12 | 61 | — | — | — | — | 13.87 | — | — | — | |
| DUET-CLIPVisual Features=CLIP-ViT/162023.04 | 72.8 | — | — | 63.4 | — | 69.3 | 60.6 | — | — | — | — | — | |
| ACME2024.12 | 72.8 | — | 3.12 | 62.3 | — | — | — | — | — | — | — | — | |
| VLN-SIGVisual Features=CLIP-ViT/162023.04 | 72.2 | — | — | 63.7 | — | 70.2 | 61.4 | — | — | — | — | — | |
| DUETMemory=Top. Map, Exploit=X2023.03 | 72 | — | 3.31 | 60 | — | — | — | — | 13.94 | — | — | — | |
| Meta-ExploreMemory=Top. Map, Exploit=local goal2023.03 | 72 | — | 3.22 | 62 | — | — | — | — | 13.09 | — | — | — | |
| DUET2023.03 | 72 | — | 3.31 | 60 | — | — | — | — | 13.94 | — | — | — | |
| DUET2024.04 | 72 | — | 3.31 | 60 | — | — | — | 81 | — | — | — | — | |
| DUET2023.12 | 72 | — | 3.31 | 60 | — | — | — | — | 13.94 | — | — | — | |
| Meta-Explore2023.12 | 72 | — | 3.22 | 62 | — | — | — | — | 13.09 | — | — | — | |
| AZHP2023.12 | 72 | — | 3.15 | 61 | — | — | — | — | 14.05 | — | — | — | |
| VLN-SIG2023.12 | 72 | — | — | 62 | — | — | — | — | — | — | — | — | |
| FDA2024.12 | 72 | — | 3.4 | 64 | — | — | — | — | — | — | — | — | |
| DUET2024.12 | 72 | — | 3.3 | 60 | — | — | — | — | — | — | — | — | |
| DUETSetting=Cross-modal backbone, Year=20222026.01 | 72 | — | 3.31 | 60 | — | — | — | 81 | 13.94 | — | — | — | |
| DUET2026.02 | 72 | — | 3.31 | 60 | — | — | — | 81 | 13.94 | — | — | — | |
| DUETSettings=Pretrain2026.02 | 72 | — | 3.31 | 60 | — | — | — | 81 | — | — | — | — | |
| DUET2026.03 | 72 | — | 3.31 | 60 | — | — | — | 81 | — | — | — | — | |
| KERM2026.03 | 72 | — | 3.22 | 61 | — | — | — | — | — | — | — | — | |
| ViTeC2026.03 | 72 | — | 3.15 | 62 | — | — | — | — | — | — | — | — | |
| DUET2026.04 | 72 | — | 3.31 | 60 | — | — | — | — | 13.94 | — | — | — | |
| DUET2026.04 | 72 | — | 3.31 | 60 | — | — | — | — | 13.94 | — | — | — | |
| DUET+SDB2026.04 | 72 | — | 3.26 | 62 | — | — | — | — | 13.49 | — | — | — | |
| NavGPT-2 (FlanT5-5B)Backbone=FlanT5-5B2025.08 | 72 | — | 3.13 | 61 | — | — | — | 81 | — | — | — | — | |
| DUET2025.08 | 72 | — | 3.31 | 60 | — | — | — | 81 | — | — | — | — | |
| DUETBase Model=DUET2026.05 | 72 | — | 3.31 | 60 | — | — | — | — | 13.94 | — | — | — | |
| TentBase Model=DUET2026.05 | 72 | — | 3.42 | 60 | — | — | — | — | 13.78 | — | — | — | |
| SARBase Model=DUET2026.05 | 72 | — | 3.28 | 61 | — | — | — | — | 13.59 | — | — | — | |
| ViDABase Model=DUET2026.05 | 72 | — | 3.34 | 61 | — | — | — | — | 13.63 | — | — | — | |
| ReCAPBase Model=DUET2026.05 | 72 | — | 3.28 | 61 | — | — | — | — | 13.39 | — | — | — |