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

Singing voice synthesis and singing voice conversion have significantlyadvanced, revolutionizing musical experiences. However, the rise of "DeepfakeSongs" generated by these technologies raises concerns about authenticity.Unlike Audio DeepFake Detection (ADD), the field of song deepfake detectionlacks specialized datasets or methods for song authenticity verification. Inthis paper, we initially construct a Chinese Fake Song Detection (FSD) datasetto investigate the field of song deepfake detection. The fake songs in the FSDdataset are generated by five state-of-the-art singing voice synthesis andsinging voice conversion methods. Our initial experiments on FSD revealed theineffectiveness of existing speech-trained ADD models for the task of songdeepFake detection. Thus, we employ the FSD dataset for the training of ADDmodels. We subsequently evaluate these models under two scenarios: one with theoriginal songs and another with separated vocal tracks. Experiment results showthat song-trained ADD models exhibit a 38.58% reduction in average equal errorrate compared to speech-trained ADD models on the FSD test set.

FSD: An Initial Chinese Dataset for Fake Song Detection