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 from single images is a challenging problem that istypically solved by supervised learning. Unfortunately, labeled training datadoes not yet exist for many human activities since 3D annotation requiresdedicated motion capture systems. Therefore, we propose an unsupervisedapproach that learns to predict a 3D human pose from a single image while onlybeing trained with 2D pose data, which can be crowd-sourced and is alreadywidely available. To this end, we estimate the 3D pose that is most likely overrandom projections, with the likelihood estimated using normalizing flows on 2Dposes. While previous work requires strong priors on camera rotations in thetraining data set, we learn the distribution of camera angles whichsignificantly improves the performance. Another part of our contribution is tostabilize training with normalizing flows on high-dimensional 3D pose data byfirst projecting the 2D poses to a linear subspace. We outperform thestate-of-the-art unsupervised human pose estimation methods on the benchmarkdatasets Human3.6M and MPI-INF-3DHP in many metrics.

ElePose: Unsupervised 3D Human Pose Estimation by Predicting Camera Elevation and Learning Normalizing Flows on 2D Poses