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

Automatic speaker verification (ASV) technology is recently finding its wayto end-user applications for secure access to personal data, smart services orphysical facilities. Similar to other biometric technologies, speakerverification is vulnerable to spoofing attacks where an attacker masquerades asa particular target speaker via impersonation, replay, text-to-speech (TTS) orvoice conversion (VC) techniques to gain illegitimate access to the system. Wefocus on TTS and VC that represent the most flexible, high-end spoofingattacks. Most of the prior studies on synthesized or converted speech detectionreport their findings using high-quality clean recordings. Meanwhile, theperformance of spoofing detectors in the presence of additive noise, animportant consideration in practical ASV implementations, remains largelyunknown. To this end, we analyze the suitability of state-of-the-art syntheticspeech detectors under additive noise with a special focus on front-endfeatures. Our comparison includes eight acoustic feature sets, five related tospectral magnitude and three to spectral phase information. Our extensiveexperiments on ASVSpoof 2015 corpus reveal several important findings. Firstly,all the countermeasures break down even at relatively high signal-to-noiseratios (SNRs) and fail to generalize to noisy conditions. Secondly, speechenhancement is not found helpful. Thirdly, GMM back-end generally outperformsthe more involved i-vector back-end. Fourthly, concerning the comparedfeatures, the Mel-frequency cepstral coefficients (MFCCs) and subband spectralcentroid magnitude coefficients (SCMCs) perform the best on average though thewinner method depends on SNR and noise type. Finally, a study with two scorefusion strategies shows that combining different feature based systems improvesrecognition accuracy for known and unknown attacks in both clean and noisyconditions.

Spoofing Detection Goes Noisy: An Analysis of Synthetic Speech Detection in the Presence of Additive Noise