7,672 human written natural language navigation instructions for routes in OpenStreetMap with a focus on visual landmarks. Validated in Street View.
Variants: map2seq
This dataset is used in 1 benchmark:
| Task | Model | Paper | Date |
|---|---|---|---|
| Vision and Language Navigation | FLAME | FLAME: Learning to Navigate with … | 2024-08-20 |
| Vision and Language Navigation | ORAR + junction type + heading delta | Analyzing Generalization of Vision and … | 2022-03-25 |
| Vision and Language Navigation | ORAR | Analyzing Generalization of Vision and … | 2022-03-25 |
| Vision and Language Navigation | Gated Attention | Analyzing Generalization of Vision and … | 2022-03-25 |
| Vision and Language Navigation | Rconcat | Analyzing Generalization of Vision and … | 2022-03-25 |
Recent papers with results on this dataset: