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

A robot in a human-centric environment needs to account for the human'sintent and future motion in its task and motion planning to ensure safe andeffective operation. This requires symbolic reasoning about probable futureactions and the ability to tie these actions to specific locations in thephysical environment. While one can train behavioral models capable ofpredicting human motion from past activities, this approach requires largeamounts of data to achieve acceptable long-horizon predictions. Moreimportantly, the resulting models are constrained to specific data formats andmodalities. Moreover, connecting predictions from such models to theenvironment at hand to ensure the applicability of these predictions is anunsolved problem. We present a system that utilizes a Large Language Model(LLM) to infer a human's next actions from a range of modalities withoutfine-tuning. A novel aspect of our system that is critical to roboticsapplications is that it links the predicted actions to specific locations in asemantic map of the environment. Our method leverages the fact that LLMs,trained on a vast corpus of text describing typical human behaviors, encodesubstantial world knowledge, including probable sequences of human actions andactivities. We demonstrate how these localized activity predictions can beincorporated in a human-aware task planner for an assistive robot to reduce theoccurrences of undesirable human-robot interactions by 29.2% on average.

GG-LLM: Geometrically Grounding Large Language Models for Zero-shot Human Activity Forecasting in Human-Aware Task Planning