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

A triangular mesh is one of the most popular 3D data representations. Assuch, the deployment of deep neural networks for mesh processing is widelyspread and is increasingly attracting more attention. However, neural networksare prone to adversarial attacks, where carefully crafted inputs impair themodel's functionality. The need to explore these vulnerabilities is afundamental factor in the future development of 3D-based applications.Recently, mesh attacks were studied on the semantic level, where classifiersare misled to produce wrong predictions. Nevertheless, mesh surfaces possesscomplex geometric attributes beyond their semantic meaning, and their analysisoften includes the need to encode and reconstruct the geometry of the shape.  We propose a novel framework for a geometric adversarial attack on a 3D meshautoencoder. In this setting, an adversarial input mesh deceives theautoencoder by forcing it to reconstruct a different geometric shape at itsoutput. The malicious input is produced by perturbing a clean shape in thespectral domain. Our method leverages the spectral decomposition of the meshalong with additional mesh-related properties to obtain visually credibleresults that consider the delicacy of surface distortions. Our code is publiclyavailable at https://github.com/StolikTomer/SAGA.

SAGA: Spectral Adversarial Geometric Attack on 3D Meshes