Vision-and-Language Navigation on map2seq Unseen (test)
46.1nDTWbest merged
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| best merged2022.03 | 46.1 | 33.5 | |
| best non-merged2022.03 | 42.2 | 30.3 | |
| ORARImage features=ResNet 4th-to-last2022.03 | 42.2 | 30.3 | |
| ORARImage features=ResNet pre-final2022.03 | 34.5 | 24.6 | |
| ORARImage features=ResNet 4th-to-last, Heading delta=false2022.03 | 33.6 | 23.6 | |
| ORARImage features=ResNet 4th-to-last, Junction type=false2022.03 | 11.8 | 7.1 | |
| ORARImage features=ResNet 4th-to-last, Heading delta=false, Junction type=false2022.03 | 8.2 | 4.7 | |
| VLN Transformer2022.03 | 6.1 | 3.5 | |
| GA2022.03 | 4.1 | 1.7 | |
| RConcat2022.03 | 3.8 | 2.1 |