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

One of the major challenges in AI is teaching machines to precisely respondand utilize environmental functionalities, thereby achieving the affordanceawareness that humans possess. Despite its importance, the field has beenlagging in terms of learning, especially in 3D, as annotating affordanceaccompanies a laborious process due to the numerous variations of human-objectinteraction. The low availability of affordance data limits the learning interms of generalization for object categories, and also simplifies therepresentation of affordance, capturing only a fraction of the affordance. Toovercome these challenges, we propose a novel, self-supervised method togenerate the 3D affordance examples given only a 3D object, without any manualannotations. The method starts by capturing the 3D object into images andcreating 2D affordance images by inserting humans into the image via inpaintingdiffusion models, where we present the Adaptive Mask algorithm to enable humaninsertion without altering the original details of the object. The methodconsequently lifts inserted humans back to 3D to create 3D human-object pairs,where the depth ambiguity is resolved within a depth optimization frameworkthat utilizes pre-generated human postures from multiple viewpoints. We alsoprovide a novel affordance representation defined on relative orientations andproximity between dense human and object points, that can be easily aggregatedfrom any 3D HOI datasets. The proposed representation serves as a primitivethat can be manifested to conventional affordance representations via simpletransformations, ranging from physically exerted affordances to nonphysicalones. We demonstrate the efficacy of our method and representation bygenerating the 3D affordance samples and deriving high-quality affordanceexamples from the representation, including contact, orientation, and spatialoccupancies.

Zero-Shot Learning for the Primitives of 3D Affordance in General Objects