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

In collaborative human-robot manipulation, a robot must predict human intentsand adapt its actions accordingly to smoothly execute tasks. However, thehuman's intent in turn depends on actions the robot takes, creating achicken-or-egg problem. Prior methods ignore such inter-dependency and insteadtrain marginal intent prediction models independent of robot actions. This isbecause training conditional models is hard given a lack of paired human-robotinteraction datasets.  Can we instead leverage large-scale human-human interaction data that is moreeasily accessible? Our key insight is to exploit a correspondence between humanand robot actions that enables transfer learning from human-human tohuman-robot data. We propose a novel architecture, InteRACT, that pre-trains aconditional intent prediction model on large human-human datasets andfine-tunes on a small human-robot dataset. We evaluate on a set of real-worldcollaborative human-robot manipulation tasks and show that our conditionalmodel improves over various marginal baselines. We also introduce newtechniques to tele-operate a 7-DoF robot arm and collect a diverse range ofhuman-robot collaborative manipulation data, which we open-source.

InteRACT: Transformer Models for Human Intent Prediction Conditioned on Robot Actions