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 have been shown to be vulnerable to small perturbationsof their inputs, known as adversarial attacks. In this paper, we investigatethe vulnerability of Neural Machine Translation (NMT) models to adversarialattacks and propose a new attack algorithm called TransFool. To fool NMTmodels, TransFool builds on a multi-term optimization problem and a gradientprojection step. By integrating the embedding representation of a languagemodel, we generate fluent adversarial examples in the source language thatmaintain a high level of semantic similarity with the clean samples.Experimental results demonstrate that, for different translation tasks and NMTarchitectures, our white-box attack can severely degrade the translationquality while the semantic similarity between the original and the adversarialsentences stays high. Moreover, we show that TransFool is transferable tounknown target models. Finally, based on automatic and human evaluations,TransFool leads to improvement in terms of success rate, semantic similarity,and fluency compared to the existing attacks both in white-box and black-boxsettings. Thus, TransFool permits us to better characterize the vulnerabilityof NMT models and outlines the necessity to design strong defense mechanismsand more robust NMT systems for real-life applications.

TransFool: An Adversarial Attack against Neural Machine Translation Models