Vision-Language Navigation on AerialVLN S (val seen)
70.3Navigation Error (NE)Grid-based VS
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Grid-based VSS.RGB=false, Depth=false, Pano.=true, Odo.=true2025.12 | 70.3 | 20.8 | 33.4 | 10.2 | |
| LookasideVLNCategory=Zero-Shot, LLM=Qwen2.5-VL-72B, Method Environment Type=Aerial2026.04 | 77.1 | 14.7 | 31.2 | 5.4 | |
| OursS.RGB=true, Depth=false, Pano.=false, Odo.=false2025.12 | 79.6 | 11.4 | 37.7 | 6.3 | |
| CityNavAgentS.RGB=false, Depth=true, Pano.=true, Odo.=true2025.12 | 80.8 | 13.9 | 30.2 | 5.1 | |
| CityNavAgentCategory=Zero-Shot, LLM=GPT-4V, Method Environment Type=Aerial2026.04 | 80.8 | 13.9 | 30.2 | 5.1 | |
| LAGS.RGB=true, Depth=true, Pano.=false, Odo.=false2025.12 | 90.2 | 7.2 | 15.7 | 2.4 | |
| LAGCategory=Learning-Based, LLM=-2026.04 | 90.2 | 7.2 | 15.7 | 2.4 | |
| SMTRS.RGB=true, Depth=true, Pano.=false, Odo.=false2025.12 | 96.3 | 12.6 | 31.6 | 2.7 | |
| STMRCategory=Zero-Shot, LLM=GPT-4o, Method Environment Type=Aerial2026.04 | 96.3 | 12.6 | 31.6 | — | |
| NavidS.RGB=true, Depth=false, Pano.=false, Odo.=false2025.12 | 105.1 | 6.8 | 15.5 | 1.1 | |
| Random SamplingS.RGB=false, Depth=false, Pano.=false, Odo.=false2025.12 | 109.6 | 0 | 0 | 0 | |
| RandomCategory=Statistical, LLM=-2026.04 | 109.6 | 0 | 0 | 0 | |
| CMAS.RGB=true, Depth=true, Pano.=false, Odo.=false2025.12 | 121 | 3 | 23.2 | 0.6 | |
| CMACategory=Learning-Based, LLM=-2026.04 | 121 | 3 | 23.2 | 0.6 | |
| MapGPTS.RGB=true, Depth=false, Pano.=false, Odo.=false2025.12 | 124.9 | 2.1 | 4.7 | 0.8 | |
| MapGPTCategory=Zero-Shot, LLM=GPT-4V, Method Environment Type=Indoor2026.04 | 124.9 | 2.1 | 4.7 | 0.8 | |
| OpenflyS.RGB=true, Depth=false, Pano.=false, Odo.=false2025.12 | 127.2 | 8.1 | 21.8 | 1.6 | |
| Seq2SeqS.RGB=true, Depth=true, Pano.=false, Odo.=false2025.12 | 146 | 4.8 | 19.8 | 1.6 | |
| Seq2SeqCategory=Learning-Based, LLM=-2026.04 | 146 | 4.8 | 19.8 | 1.6 | |
| NavGPTCategory=Zero-Shot, LLM=GPT-4, Method Environment Type=Indoor2026.04 | 163.5 | 0 | 0 | 0 | |
| Action SamplingS.RGB=false, Depth=false, Pano.=false, Odo.=false2025.12 | 213.8 | 0.9 | 5.7 | 0.3 | |
| Action SamplingCategory=Statistical, LLM=-2026.04 | 213.8 | 0.9 | 5.7 | 0.3 | |
| LingUNetS.RGB=true, Depth=true, Pano.=false, Odo.=false2025.12 | 383.8 | 0.6 | 6.9 | 0.2 |