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

Training accurate 3D human pose estimators requires large amount of 3Dground-truth data which is costly to collect. Various weakly or self supervisedpose estimation methods have been proposed due to lack of 3D data.Nevertheless, these methods, in addition to 2D ground-truth poses, requireeither additional supervision in various forms (e.g. unpaired 3D ground truthdata, a small subset of labels) or the camera parameters in multiview settings.To address these problems, we present EpipolarPose, a self-supervised learningmethod for 3D human pose estimation, which does not need any 3D ground-truthdata or camera extrinsics. During training, EpipolarPose estimates 2D posesfrom multi-view images, and then, utilizes epipolar geometry to obtain a 3Dpose and camera geometry which are subsequently used to train a 3D poseestimator. We demonstrate the effectiveness of our approach on standardbenchmark datasets i.e. Human3.6M and MPI-INF-3DHP where we set the newstate-of-the-art among weakly/self-supervised methods. Furthermore, we proposea new performance measure Pose Structure Score (PSS) which is a scaleinvariant, structure aware measure to evaluate the structural plausibility of apose with respect to its ground truth. Code and pretrained models are availableat https://github.com/mkocabas/EpipolarPose

Self-Supervised Learning of 3D Human Pose using Multi-view Geometry