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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

We present the first method to perform automatic 3D pose, shape and texturecapture of animals from images acquired in-the-wild. In particular, we focus onthe problem of capturing 3D information about Grevy's zebras from a collectionof images. The Grevy's zebra is one of the most endangered species in Africa,with only a few thousand individuals left. Capturing the shape and pose ofthese animals can provide biologists and conservationists with informationabout animal health and behavior. In contrast to research on human pose, shapeand texture estimation, training data for endangered species is limited, theanimals are in complex natural scenes with occlusion, they are naturallycamouflaged, travel in herds, and look similar to each other. To overcome thesechallenges, we integrate the recent SMAL animal model into a network-basedregression pipeline, which we train end-to-end on synthetically generatedimages with pose, shape, and background variation. Going beyondstate-of-the-art methods for human shape and pose estimation, our method learnsa shape space for zebras during training. Learning such a shape space fromimages using only a photometric loss is novel, and the approach can be used tolearn shape in other settings with limited 3D supervision. Moreover, we couple3D pose and shape prediction with the task of texture synthesis, obtaining afull texture map of the animal from a single image. We show that the predictedtexture map allows a novel per-instance unsupervised optimization over thenetwork features. This method, SMALST (SMAL with learned Shape and Texture)goes beyond previous work, which assumed manual keypoints and/or segmentation,to regress directly from pixels to 3D animal shape, pose and texture. Code anddata are available at https://github.com/silviazuffi/smalst.

Three-D Safari: Learning to Estimate Zebra Pose, Shape, and Texture from Images "In the Wild"