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

With the recent advancements of deep learning technologies, the performanceof voice conversion (VC) in terms of quality and similarity has beensignificantly improved. However, heavy computations are generally required fordeep-learning-based VC systems, which can cause notable latency and thusconfine their deployments in real-world applications. Therefore, increasingonline computation efficiency has become an important task. In this study, wepropose a novel mixture-of-experts (MoE) based VC system. The MoE model uses agating mechanism to specify optimal weights to feature maps to increase VCperformance. In addition, assigning sparse constraints on the gating mechanismcan accelerate online computation by skipping the convolution process byzeroing out redundant feature maps. Experimental results show that byspecifying suitable sparse constraints, we can effectively increase the onlinecomputation efficiency with a notable 70% FLOPs (floating-point operations persecond) reduction while improving the VC performance in both objectiveevaluations and human listening tests.

MoEVC: A Mixture-of-experts Voice Conversion System with Sparse Gating Mechanism for Accelerating Online Computation