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

Recent research has revealed that the security of deep neural networks thatdirectly process 3D point clouds to classify objects can be threatened byadversarial samples. Although existing adversarial attack methods achieve highsuccess rates, they do not restrict the point modifications enough to preservethe point cloud appearance. To overcome this shortcoming, two constraints areproposed. These include applying hard boundary constraints on the number ofmodified points and on the point perturbation norms. Due to the restrictivenature of the problem, the search space contains many local maxima. Theproposed method addresses this issue by using a high step-size at the beginningof the algorithm to search the main surface of the point cloud fast andeffectively. Then, in order to converge to the desired output, the step-size isgradually decreased. To evaluate the performance of the proposed method, it isrun on the ModelNet40 and ScanObjectNN datasets by employing thestate-of-the-art point cloud classification models; including PointNet,PointNet++, and DGCNN. The obtained results show that it can perform successfulattacks and achieve state-of-the-art results by only a limited number of pointmodifications while preserving the appearance of the point cloud. Moreover, dueto the effective search algorithm, it can perform successful attacks in just afew steps. Additionally, the proposed step-size scheduling algorithm shows animprovement of up to $14.5\%$ when adopted by other methods as well. Theproposed method also performs effectively against popular defense methods.

Adversarial Attack by Limited Point Cloud Surface Modifications