Vision-and-Language Navigation on Average cross-dataset (unseen)
29.3Success Rate (SR)BABYWALK
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| BABYWALKTraining Dataset=R4R, Data Augmentation=false2020.05 | 29.3 | 46 | 17.3 | |
| BABYWALKTraining Dataset=R4R, Data Augmentation=true2020.05 | 27.6 | 47.9 | 17.5 | |
| FASTTraining Dataset=R4R, Data Augmentation=true, Reimplemented=true2020.05 | 25 | 31.7 | 9.4 | |
| RCM(FIDELITY)Training Dataset=R4R, Data Augmentation=true2020.05 | 23.5 | 35.7 | 10.8 | |
| RCM(GOAL)Training Dataset=R4R, Data Augmentation=true2020.05 | 22.7 | 34.5 | 10.9 | |
| SFTraining Dataset=R4R, Data Augmentation=true2020.05 | 19.9 | 22.4 | 8.3 | |
| REGRETFULTraining Dataset=R4R, Data Augmentation=true, Reimplemented=true2020.05 | 19.8 | 31.2 | 8.8 | |
| SEQ2SEQTraining Dataset=R4R, Data Augmentation=false2020.05 | 17.1 | 19.9 | 6.6 |