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

Generative adversarial network (GAN)-based neural vocoders have been widelyused in audio synthesis tasks due to their high generation quality, efficientinference, and small computation footprint. However, it is still challenging totrain a universal vocoder which can generalize well to out-of-domain (OOD)scenarios, such as unseen speaking styles, non-speech vocalization, singing,and musical pieces. In this work, we propose SnakeGAN, a GAN-based universalvocoder, which can synthesize high-fidelity audio in various OOD scenarios.SnakeGAN takes a coarse-grained signal generated by a differentiable digitalsignal processing (DDSP) model as prior knowledge, aiming at recoveringhigh-fidelity waveform from a Mel-spectrogram. We introduce periodicnonlinearities through the Snake activation function and anti-aliasedrepresentation into the generator, which further brings the desired inductivebias for audio synthesis and significantly improves the extrapolation capacityfor universal vocoding in unseen scenarios. To validate the effectiveness ofour proposed method, we train SnakeGAN with only speech data and evaluate itsperformance for various OOD distributions with both subjective and objectivemetrics. Experimental results show that SnakeGAN significantly outperforms thecompared approaches and can generate high-fidelity audio samples includingunseen speakers with unseen styles, singing voices, instrumental pieces, andnonverbal vocalization.

SnakeGAN: A Universal Vocoder Leveraging DDSP Prior Knowledge and Periodic Inductive Bias