MARVAL

Reported on 6 benchmarks across 1 task · 1 paper · 1 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Robots6 results

Vision and Language NavigationonRxR
ndtw· uses extra data· 2022-10-06
66.76
SOTA
A New Path: Scaling Vision-and-Language Navigation with Synthetic Instructions and Imitation Learning arXiv:2210.03112
Vision and Language NavigationonVLN Challenge
error
4.18
best: 1.61 (human)
Vision and Language NavigationonVLN Challenge
length
10.22
best: 1257.38 (Speaker-Follower)
Vision and Language NavigationonVLN Challenge
oracle success
0.67
best: 1 (FOAM-Beam Search)
Vision and Language NavigationonVLN Challenge
spl
0.58
best: 0.76 (human)
Vision and Language NavigationonVLN Challenge
success
0.62
best: 0.86 (human)