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

Non-Intrusive speech quality assessment (NISQA) has gained significantattention for predicting the mean opinion score (MOS) of speech withoutrequiring the reference speech. In practical NISQA scenarios, untrustedthird-party resources are often employed during deep neural network training toreduce costs. However, it would introduce a potential security vulnerability asspecially designed untrusted resources can launch backdoor attacks againstNISQA systems. Existing backdoor attacks primarily focus on classificationtasks and are not directly applicable to NISQA which is a regression task. Inthis paper, we propose a novel backdoor attack on NISQA tasks, leveragingpresence events as triggers to achieving highly stealthy attacks. To evaluatethe effectiveness of our proposed approach, we conducted experiments on fourbenchmark datasets and employed two state-of-the-art NISQA models. The resultsdemonstrate that the proposed backdoor attack achieved an average attacksuccess rate of up to 99% with a poisoning rate of only 3%.

BadSQA: Stealthy Backdoor Attacks Using Presence Events as Triggers in Non-Intrusive Speech Quality Assessment