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

We present a new local descriptor for 3D shapes, directly applicable to awide range of shape analysis problems such as point correspondences, semanticsegmentation, affordance prediction, and shape-to-scan matching. The descriptoris produced by a convolutional network that is trained to embed geometricallyand semantically similar points close to one another in descriptor space. Thenetwork processes surface neighborhoods around points on a shape that arecaptured at multiple scales by a succession of progressively zoomed out views,taken from carefully selected camera positions. We leverage two extremely largesources of data to train our network. First, since our network processesrendered views in the form of 2D images, we repurpose architectures pre-trainedon massive image datasets. Second, we automatically generate a synthetic densepoint correspondence dataset by non-rigid alignment of corresponding shapeparts in a large collection of segmented 3D models. As a result of these designchoices, our network effectively encodes multi-scale local context andfine-grained surface detail. Our network can be trained to produce eithercategory-specific descriptors or more generic descriptors by learning frommultiple shape categories. Once trained, at test time, the network extractslocal descriptors for shapes without requiring any part segmentation as input.Our method can produce effective local descriptors even for shapes whosecategory is unknown or different from the ones used while training. Wedemonstrate through several experiments that our learned local descriptors aremore discriminative compared to state of the art alternatives, and areeffective in a variety of shape analysis applications.

Learning Local Shape Descriptors from Part Correspondences With Multi-view Convolutional Networks