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

Deep neural networks (DNNs) are widely applied for nowadays 3D surfacereconstruction tasks and such methods can be further divided into twocategories, which respectively warp templates explicitly by moving vertices orrepresent 3D surfaces implicitly as signed or unsigned distance functions.Taking advantage of both advanced explicit learning process and powerfulrepresentation ability of implicit functions, we propose a novel 3Drepresentation method, Neural Vector Fields (NVF). It not only adopts theexplicit learning process to manipulate meshes directly, but also leverages theimplicit representation of unsigned distance functions (UDFs) to break thebarriers in resolution and topology. Specifically, our method first predictsthe displacements from queries towards the surface and models the shapes as\textit{Vector Fields}. Rather than relying on network differentiation toobtain direction fields as most existing UDF-based methods, the produced vectorfields encode the distance and direction fields both and mitigate the ambiguityat "ridge" points, such that the calculation of direction fields isstraightforward and differentiation-free. The differentiation-freecharacteristic enables us to further learn a shape codebook via VectorQuantization, which encodes the cross-object priors, accelerates the trainingprocedure, and boosts model generalization on cross-category reconstruction.The extensive experiments on surface reconstruction benchmarks indicate thatour method outperforms those state-of-the-art methods in different evaluationscenarios including watertight vs non-watertight shapes, category-specific vscategory-agnostic reconstruction, category-unseen reconstruction, andcross-domain reconstruction. Our code is released athttps://github.com/Wi-sc/NVF.

Neural Vector Fields: Implicit Representation by Explicit Learning