Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Beyond the Nav-Graph: Vision-and-Language Navigation in Continuous Environments

About

We develop a language-guided navigation task set in a continuous 3D environment where agents must execute low-level actions to follow natural language navigation directions. By being situated in continuous environments, this setting lifts a number of assumptions implicit in prior work that represents environments as a sparse graph of panoramas with edges corresponding to navigability. Specifically, our setting drops the presumptions of known environment topologies, short-range oracle navigation, and perfect agent localization. To contextualize this new task, we develop models that mirror many of the advances made in prior settings as well as single-modality baselines. While some of these techniques transfer, we find significantly lower absolute performance in the continuous setting -- suggesting that performance in prior `navigation-graph' settings may be inflated by the strong implicit assumptions.

Jacob Krantz, Erik Wijmans, Arjun Majumdar, Dhruv Batra, Stefan Lee• 2020

Related benchmarks

TaskDatasetResultRank
Vision-Language NavigationR2R-CE (val-unseen)
Success Rate (SR)41
433
Vision-and-Language NavigationR2R (val unseen)
Success Rate (SR)25
344
Vision-Language NavigationRxR-CE (val-unseen)
SR26.5
280
Vision-and-Language NavigationR2R-CE (test-unseen)
SR28
63
Vision-and-Language NavigationR2R-CE (val-seen)
SR37
49
Vision-and-Language NavigationR2R-CE unseen continuous (val)
SR41
35
Vertical PerceptionNavSpace
Navigation Error (NE)7.88
30
Precise MovementNavSpace
Navigation Error (NE)6.85
27
Vision-Language NavigationRxR (val-unseen)
Navigation Error (NE)12.1
25
Vision-Language NavigationHA-VLN Unseen (val)
NE7.34
23
Showing 10 of 32 rows

Other info

Follow for update