Vision-Language Navigation on R4R unseen v1.0 (val)
11.6Path LengthRCM(GOAL)+
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| RCM(GOAL)+pre-trained with data augmentation=true2020.05 | 11.6 | 8.5 | 23.6 | 17.5 | 35.8 | 29.8 | 10.8 | |
| BABYWALK +pre-trained with data augmentation=true2020.05 | 17.7 | 8.2 | 24.3 | 12.9 | 48.6 | 39.4 | 15.1 | |
| BABY WALK2020.05 | 18.1 | 8.4 | 24.3 | 12.8 | 48.6 | 39 | 15.1 | |
| SEQ2SEQ2020.05 | 25.2 | 8.7 | 24.2 | 13.7 | 25.8 | 22.9 | 9.3 | |
| RCM(FIDELITY)+pre-trained with data augmentation=true2020.05 | 25.7 | 8.4 | 25.4 | 10.6 | 34.8 | 26.5 | 11.1 | |
| SF+pre-trained with data augmentation=true2020.05 | 33 | 8.6 | 22.4 | 9.3 | 21.4 | 20.6 | 7.5 |