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

In this paper, we propose a novel self-distillation method for fake speechdetection (FSD), which can significantly improve the performance of FSD withoutincreasing the model complexity. For FSD, some fine-grained information is veryimportant, such as spectrogram defects, mute segments, and so on, which areoften perceived by shallow networks. However, shallow networks have much noise,which can not capture this very well. To address this problem, we propose usingthe deepest network instruct shallow network for enhancing shallow networks.Specifically, the networks of FSD are divided into several segments, thedeepest network being used as the teacher model, and all shallow networksbecome multiple student models by adding classifiers. Meanwhile, thedistillation path between the deepest network feature and shallow networkfeatures is used to reduce the feature difference. A series of experimentalresults on the ASVspoof 2019 LA and PA datasets show the effectiveness of theproposed method, with significant improvements compared to the baseline.

Learning From Yourself: A Self-Distillation Method for Fake Speech Detection