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

State-of-the-art deep neural networks have been shown to be extremelypowerful in a variety of perceptual tasks like semantic segmentation. However,these networks are vulnerable to adversarial perturbations of the input whichare imperceptible for humans but lead to incorrect predictions. Treating imagesegmentation as a sum of pixel-wise classifications, adversarial attacksdeveloped for classification models were shown to be applicable to segmentationmodels as well. In this work, we present simple uncertainty-based weightingschemes for the loss functions of such attacks that (i) put higher weights onpixel classifications which can more easily perturbed and (ii) zero-out thepixel-wise losses corresponding to those pixels that are already confidentlymisclassified. The weighting schemes can be easily integrated into the lossfunction of a range of well-known adversarial attackers with minimal additionalcomputational overhead, but lead to significant improved perturbationperformance, as we demonstrate in our empirical analysis on several datasetsand models.

Uncertainty-weighted Loss Functions for Improved Adversarial Attacks on Semantic Segmentation