Vision-and-Language Navigation on R2R-CE (val-seen)
73SRCLASH
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| CLASHType=III, Panoramic Visual Input=true, Additional Data Beyond VLN=false2025.12 | 73 | 12.4 | 3.2 | — | 80 | 65 | |
| BEVBert + FEEDTTALearning Paradigm=With Feedback, Base Model=BEVBert, Method Adjustment=FEEDTTA2026.05 | 73 | 13.54 | 3.08 | — | 79 | 63 | |
| IDEALearning Paradigm=Self-supervised, Base Model=BEVBert, Method Adjustment=IDEA2026.05 | 73 | 12.27 | 3.04 | — | 78 | 64 | |
| IDEAPolicy=BEVBert2026.05 | 73 | 12.27 | 3.04 | — | 78 | 64 | |
| BEVBert + VIL (Ours)Viewpoint setting=standard viewpoint2025.07 | 71.1 | — | 3.16 | — | — | 63 | |
| Safe-VLNParadigm=Explicit Map-based2026.01 | 71 | 13.71 | 3.35 | — | 79 | 60 | |
| BEVBert + ATENALearning Paradigm=With Feedback, Base Model=BEVBert, Method Adjustment=ATENA2026.05 | 71 | 11.31 | 3.24 | — | 75 | 64 | |
| BEVBert + ReCAPLearning Paradigm=Self-supervised, Base Model=BEVBert, Method Adjustment=ReCAP2026.05 | 71 | 12.4 | 3.31 | — | 76 | 63 | |
| ViDAPolicy=BEVBert2026.05 | 71 | 12.61 | 3.31 | — | 76 | 62 | |
| ReCAPPolicy=BEVBert2026.05 | 71 | 12.4 | 3.31 | — | 76 | 63 | |
| BEVBertViewpoint setting=standard viewpoint2025.07 | 70.9 | — | 3.24 | — | — | 62.8 | |
| UnitedVLNType=I, Panoramic Visual Input=true, Additional Data Beyond VLN=false2025.12 | 70 | — | 3.3 | — | 78 | 62 | |
| TentPolicy=BEVBert2026.05 | 70 | 12.74 | 3.35 | — | 76 | 62 | |
| SARPolicy=BEVBert2026.05 | 70 | 12.53 | 3.28 | — | 76 | 63 | |
| HNR2024.10 | 69 | 11.79 | 3.67 | — | 76 | 61 | |
| HNRType=I, Panoramic Visual Input=true, Additional Data Beyond VLN=false2025.12 | 69 | 11.79 | 3.67 | — | 76 | 61 | |
| BEVBert + RLCFLearning Paradigm=With Feedback, Base Model=BEVBert, Method Adjustment=RLCF2026.05 | 69 | 13.74 | 3.56 | — | 74 | 61 | |
| BEVBert + FSTTALearning Paradigm=Self-supervised, Base Model=BEVBert, Method Adjustment=FSTTA2026.05 | 69 | 14.07 | 4.11 | — | 74 | 60 | |
| FSTTAPolicy=BEVBert2026.05 | 69 | 14.07 | 4.11 | — | 74 | 60 | |
| HNRTraining paradigm=Scheduled Sampling, Observation view=panoramic-view, Φ-Nav application=false2026.07 | 69 | 11.79 | 3.67 | — | 76 | 61 | |
| TEN2026.06 | 68.8 | — | 3.75 | — | 74.6 | 60.2 | |
| ETPNav w/ Φ-NavTraining paradigm=Scheduled Sampling, Observation view=panoramic-view, Φ-Nav application=true2026.07 | 68.63 | 11.29 | 3.38 | — | 75.06 | 61.41 | |
| ENP-ETPNav2024.10 | 68 | 11.82 | 3.9 | — | 73 | 59 | |
| OVL-MAPParadigm=Explicit Map-based2026.01 | 68 | 11.82 | 3.9 | — | 73 | 59 | |
| ENPViewpoint setting=standard viewpoint2025.07 | 68 | — | 3.9 | — | — | 59 | |
| BEVBertLearning Paradigm=With Feedback, Base Model=BEVBert2026.05 | 68 | 13.98 | 3.77 | — | 73 | 60 | |
| BEVBertLearning Paradigm=Self-supervised, Base Model=BEVBert2026.05 | 68 | 13.98 | 3.77 | — | 73 | 60 | |
| IDEALearning Paradigm=Self-supervised, Base Model=ETPNav, Method Adjustment=IDEA2026.05 | 68 | 10.87 | 3.82 | — | 73 | 62 | |
| BEVBertPolicy=BEVBert2026.05 | 68 | 13.98 | 3.77 | — | 73 | 60 | |
| IDEAPolicy=ETPNav2026.05 | 68 | 10.87 | 3.82 | — | 73 | 62 | |
| BEVBertWaypoint predictor=from [16]2026.06 | 68 | — | 3.77 | — | 73 | 60 | |
| EnergyWaypoint predictor=from [16]2026.06 | 68 | — | 3.9 | — | 73 | 59 | |
| BEVBertTraining paradigm=Scheduled Sampling, Observation view=panoramic-view, Φ-Nav application=false2026.07 | 68 | 13.98 | 3.77 | — | 73 | 60 | |
| ENP-ETPNavTraining paradigm=Scheduled Sampling, Observation view=panoramic-view, Φ-Nav application=false2026.07 | 68 | 11.82 | 3.9 | — | 73 | 59 | |
| ETPNav + VIL (Ours)Viewpoint setting=standard viewpoint2025.07 | 67.6 | — | 3.71 | — | — | 60.4 | |
| ETPNav + FeedTTALearning Paradigm=With Feedback, Base Model=ETPNav, Method Adjustment=FeedTTA2026.05 | 67 | 10.88 | 3.85 | — | 72 | 61 | |
| ETPNav + ATENALearning Paradigm=With Feedback, Base Model=ETPNav, Method Adjustment=ATENA2026.05 | 67 | 10.81 | 3.86 | — | 72 | 61 | |
| ETPNavTraining paradigm=Scheduled Sampling, Observation view=panoramic-view, Φ-Nav application=false2026.07 | 66.45 | 11.38 | 3.94 | — | 72.23 | 59.62 | |
| ETPNav2024.10 | 66 | 11.78 | 3.95 | — | 72 | 59 | |
| ETPNavParadigm=Explicit Map-based2026.01 | 66 | 11.78 | 3.95 | — | 72 | 59 | |
| DGNavParadigm=Explicit Map-based2026.01 | 66 | 10.79 | 3.78 | — | 72 | 60 | |
| ETPNavType=I, Panoramic Visual Input=true, Additional Data Beyond VLN=false2025.12 | 66 | 11.78 | 3.95 | — | 72 | 59 | |
| ETPNavLearning Paradigm=With Feedback, Base Model=ETPNav2026.05 | 66 | 11.78 | 3.95 | — | 72 | 59 | |
| ETPNav + RLCFLearning Paradigm=With Feedback, Base Model=ETPNav, Method Adjustment=RLCF2026.05 | 66 | 11.45 | 3.91 | — | 72 | 59 | |
| ETPNavLearning Paradigm=Self-supervised, Base Model=ETPNav2026.05 | 66 | 11.78 | 3.95 | — | 72 | 59 | |
| ETPNav + FSTTALearning Paradigm=Self-supervised, Base Model=ETPNav, Method Adjustment=FSTTA2026.05 | 66 | 11.35 | 3.93 | — | 72 | 59 | |
| ETPNav + ReCAPLearning Paradigm=Self-supervised, Base Model=ETPNav, Method Adjustment=ReCAP2026.05 | 66 | 11.31 | 3.92 | — | 72 | 60 | |
| ETPNavPolicy=ETPNav2026.05 | 66 | 11.78 | 3.95 | — | 72 | 59 | |
| TentPolicy=ETPNav2026.05 | 66 | 11.36 | 3.94 | — | 72 | 59 | |
| SARPolicy=ETPNav2026.05 | 66 | 11.31 | 3.89 | — | 72 | 60 | |
| ViDAPolicy=ETPNav2026.05 | 66 | 11.58 | 3.91 | — | 72 | 60 | |
| FSTTAPolicy=ETPNav2026.05 | 66 | 11.35 | 3.93 | — | 72 | 59 | |
| ReCAPPolicy=ETPNav2026.05 | 66 | 11.31 | 3.92 | — | 72 | 60 | |
| ETPNavWaypoint predictor=from [16]2026.06 | 66 | — | 3.95 | — | 72 | 59 | |
| ETPNavViewpoint setting=standard viewpoint2025.07 | 65.8 | — | 3.97 | — | — | 59.2 | |
| g3D-LFTraining paradigm=Scheduled Sampling, Observation view=panoramic-view, Φ-Nav application=false2026.07 | 63 | 11.61 | 3.72 | — | 72 | 55 | |
| StreamVLNType=II, Panoramic Visual Input=false, Additional Data Beyond VLN=true2025.12 | 62 | — | 4.53 | — | 68 | 58 | |
| Uni-NavidVLM-based=true2026.06 | 59.1 | — | 4.4 | — | 64.5 | 55.4 | |
| DREAM2024.10 | 59 | 11.6 | 4.09 | — | 66 | 48 | |
| GridMMParadigm=Explicit Map-based2026.01 | 59 | 12.69 | 4.21 | — | 69 | 51 | |
| GridMMType=I, Panoramic Visual Input=true, Additional Data Beyond VLN=false2025.12 | 59 | 12.69 | 4.21 | — | 69 | 51 | |
| DreamWalkerType=I, Panoramic Visual Input=true, Additional Data Beyond VLN=false2025.12 | 59 | 11.6 | 4.09 | — | 66 | 48 | |
| GridMMWaypoint predictor=from [16]2026.06 | 59 | — | 4.21 | — | 69 | 51 | |
| DWalkerWaypoint predictor=from [16]2026.06 | 59 | — | 5.53 | — | 66 | 48 | |
| Uni-NaVidType=II, Panoramic Visual Input=false, Additional Data Beyond VLN=true2025.12 | 58 | 10.04 | 4.45 | — | 65 | 53 | |
| NaVILAType=II, Panoramic Visual Input=false, Additional Data Beyond VLN=true2025.12 | 58 | — | 4.82 | — | 66 | 53 | |
| NaVILAVLM-based=true2026.06 | 56.4 | — | 5.05 | — | 63.8 | 51.5 | |
| CM22026.06 | 52.8 | — | 4.81 | — | 58.3 | 41.8 | |
| Sim2Sim2024.10 | 52 | 11.18 | 4.67 | — | 61 | 44 | |
| Sim2SimParadigm=End-to-End2026.01 | 52 | 11.18 | 4.67 | — | 61 | 44 | |
| Sim2SimWaypoint predictor=from [16]2026.06 | 52 | — | 4.67 | — | 61 | 44 | |
| Waypoint PredictorBase Agent=CMA2022.03 | 51 | 11.47 | 5.2 | 61 | 61 | 45 | |
| ENP-VLNOBERT2024.10 | 51 | 12.08 | 4.89 | — | 60 | 44 | |
| CWP-CMAParadigm=End-to-End2026.01 | 51 | 11.47 | 5.2 | — | 61 | 45 | |
| CWP-CMAType=I, Panoramic Visual Input=true, Additional Data Beyond VLN=false2025.12 | 51 | 11.47 | 5.2 | — | 61 | 45 | |
| Waypoint PredictorBase Agent=VLN BERT2022.03 | 50 | 12.5 | 5.02 | 58 | 59 | 44 | |
| VLNOBERT2024.10 | 50 | 12.5 | 5.02 | — | 59 | 44 | |
| CWP-RecBertParadigm=End-to-End2026.01 | 50 | 12.5 | 5.02 | — | 59 | 44 | |
| CWP-RecBERTType=I, Panoramic Visual Input=true, Additional Data Beyond VLN=false2025.12 | 50 | 12.5 | 5.02 | — | 59 | 44 | |
| VLN⟳BERTViewpoint setting=standard viewpoint2025.07 | 50 | — | 5.02 | — | — | 44 | |
| VLN⟳BERTWaypoint predictor=from [16]2026.06 | 50 | — | 5.74 | — | 59 | 44 | |
| VLN-BERTTraining paradigm=Scheduled Sampling, Observation view=panoramic-view, Φ-Nav application=false2026.07 | 50 | 12.5 | 5.02 | — | 59 | 44 | |
| WSMGMap2024.10 | 47 | 10.12 | 5.65 | — | 52 | 43 | |
| WG-MGMapParadigm=Explicit Map-based2026.01 | 47 | 10.12 | 5.65 | — | 52 | 43 | |
| Waypoint Models2022.03 | 46 | 8.54 | 5.48 | — | 53 | 43 | |
| HPN2024.10 | 46 | 8.54 | 5.48 | — | 53 | 43 | |
| HPN+DN2026.06 | 46 | — | 5.48 | — | 53 | 43 | |
| NavidVLM-based=true2026.06 | 43.5 | — | 5.62 | — | 51.5 | 38.4 | |
| CM22024.10 | 43 | 12.05 | 6.1 | — | 51 | 35 | |
| CM2Type=I, Panoramic Visual Input=false, Additional Data Beyond VLN=false2025.12 | 43 | 12.05 | 6.1 | — | 51 | 35 | |
| NaVidType=II, Panoramic Visual Input=false, Additional Data Beyond VLN=false2025.12 | 43 | 10.27 | 5.47 | — | 52 | 38 | |
| CM2Paradigm=Explicit Map-based2026.01 | 42.9 | 12.05 | 6.1 | — | 50.7 | 34.8 | |
| CMA-D-P-A w/ Φ-NavTraining paradigm=DAgger, Observation view=monocular-view, Φ-Nav application=true2026.07 | 38.8 | 9.54 | 6.99 | — | 47.18 | 35.63 | |
| CMA-D-P-ATraining paradigm=DAgger, Observation view=monocular-view, Φ-Nav application=false2026.07 | 37.17 | 9.63 | 7.11 | — | 46.13 | 34.82 | |
| VLN-CE2022.03 | 37 | 9.26 | 7.12 | 54 | 46 | 35 | |
| LAW2022.03 | 37 | — | — | 58 | 40 | — | |
| VLN-CE2024.10 | 37 | 9.26 | 7.12 | — | 46 | 35 | |
| Seq2SeqParadigm=End-to-End2026.01 | 37 | 9.26 | 7.12 | — | 46 | 35 | |
| SASRA2022.03 | 36 | 8.89 | 7.17 | 53 | — | 34 | |
| CMTP2024.10 | 36 | — | 7.1 | — | 56 | 31 |