English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Embodied instruction following is a challenging problem requiring an agent toinfer a sequence of primitive actions to achieve a goal environment state fromcomplex language and visual inputs. Action Learning From Realistic Environmentsand Directives (ALFRED) is a recently proposed benchmark for this problemconsisting of step-by-step natural language instructions to achieve subgoalswhich compose to an ultimate high-level goal. Key challenges for this taskinclude localizing target locations and navigating to them through visualinputs, and grounding language instructions to visual appearance of objects. Toaddress these challenges, in this study, we augment the agent's field of viewduring navigation subgoals with multiple viewing angles, and train the agent topredict its relative spatial relation to the target location at each timestep.We also improve language grounding by introducing a pre-trained objectdetection module to the model pipeline. Empirical studies show that ourapproach exceeds the baseline model performance.

Are We There Yet? Learning to Localize in Embodied Instruction Following