D3AdvM: A direct 3D adversarial sample attack inside mesh data

DOI10.1016/j.cagd.2022.102122OpenAlexW4285005480WikidataQ114202275 ScholiaQ114202275MaRDI QIDQ2168854

Junwei Bai, Huangxinxin Xu, Linkun Fan, Fazhi He

Publication date: 26 August 2022

Published in: Computer Aided Geometric Design (Search for Journal in Brave)

Full work available at URL: https://doi.org/10.1016/j.cagd.2022.102122

zbMATH Keywords

deep learning3D meshtransferabilityadversarial sample attack

Mathematics Subject Classification ID

Computer science (68-XX)

Related Items (1)

TPNet: a novel mesh analysis method via topology preservation and perception enhancement

Cites Work

• Least squares quasi-developable mesh approximation
• Unknots with highly knotted control polygons
• Composite \(\sqrt{2}\) subdivision surfaces
• Advancing mesh completion for digital modeling and manufacturing
• Multi-resolution 3D CNN for learning multi-scale spatial features in CAD models
• Affine invariant triangulations
• EasyMesh: an efficient method to reconstruct 3D mesh from a single image
• General triangular midpoint subdivision
• Analysis-suitable adaptive T-mesh refinement with linear complexity
• Truly selective polygonal mesh hierarchies with error control
• Approximation by polynomial splines on curved triangulations
• EDITORIAL: CONTENT-SENSITIVE METHODS FOR GEOMETRIC MODELING

This page was built for publication: D3AdvM: A direct 3D adversarial sample attack inside mesh data