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

Human pose estimation is a major computer vision problem with applicationsranging from augmented reality and video capture to surveillance and movementtracking. In the medical context, the latter may be an important biomarker forneurological impairments in infants. Whilst many methods exist, theirapplication has been limited by the need for well annotated large datasets andthe inability to generalize to humans of different shapes and bodycompositions, e.g. children and infants. In this paper we present a novelmethod for learning pose estimators for human adults and infants in anunsupervised fashion. We approach this as a learnable template matching problemfacilitated by deep feature extractors. Human-interpretable landmarks areestimated by transforming a template consisting of predefined body parts thatare characterized by 2D Gaussian distributions. Enforcing a connectivity priorguides our model to meaningful human shape representations. We demonstrate theeffectiveness of our approach on two different datasets including adults andinfants.

Unsupervised Human Pose Estimation through Transforming Shape Templates