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/

Zendy Plus

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

Zendy Tools

Zendy Open

Human motion prediction is a complex task as it involves forecastingvariables over time on a graph of connected sensors. This is especially true inthe case of few-shot learning, where we strive to forecast motion sequences forpreviously unseen actions based on only a few examples. Despite this, almostall related approaches for few-shot motion prediction do not incorporate theunderlying graph, while it is a common component in classical motionprediction. Furthermore, state-of-the-art methods for few-shot motionprediction are restricted to motion tasks with a fixed output space meaningthese tasks are all limited to the same sensor graph. In this work, we proposeto extend recent works on few-shot time-series forecasting with heterogeneousattributes with graph neural networks to introduce the first few-shot motionapproach that explicitly incorporates the spatial graph while also generalizingacross motion tasks with heterogeneous sensors. In our experiments on motiontasks with heterogeneous sensors, we demonstrate significant performanceimprovements with lifts from 10.4% up to 39.3% compared to beststate-of-the-art models. Moreover, we show that our model can perform on parwith the best approach so far when evaluating on tasks with a fixed outputspace while maintaining two magnitudes fewer parameters.

Few-shot human motion prediction for heterogeneous sensors