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

The fast increase of web services and mobile apps, which collect personaldata from users, increases the risk that their privacy may be severelycompromised. In particular, the increasing variety of spoken languageinterfaces and voice assistants empowered by the vertiginous breakthroughs inDeep Learning are prompting important concerns in the European Union topreserve speech data privacy. For instance, an attacker can record speech fromusers and impersonate them to get access to systems requiring voiceidentification. Hacking speaker profiles from users is also possible by meansof existing technology to extract speaker, linguistic (e.g., dialect) andparalinguistic features (e.g., age) from the speech signal. In order tomitigate these weaknesses, in this paper, we propose a speakerde-identification system based on adversarial training and autoencoders inorder to suppress speaker, gender, and accent information from speech.Experimental results show that combining adversarial learning and autoencodersincrease the equal error rate of a speaker verification system while preservingthe intelligibility of the anonymized spoken content.

Speaker De-identification System using Autoencoders and Adversarial Training