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/

Zendy Plus

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

Zendy Tools

Zendy Open

Non-parallel voice conversion (VC) is a technique for learning mappingsbetween source and target speeches without using a parallel corpus. Recently,cycle-consistent adversarial network (CycleGAN)-VC and CycleGAN-VC2 have shownpromising results regarding this problem and have been widely used as benchmarkmethods. However, owing to the ambiguity of the effectiveness ofCycleGAN-VC/VC2 for mel-spectrogram conversion, they are typically used formel-cepstrum conversion even when comparative methods employ mel-spectrogram asa conversion target. To address this, we examined the applicability ofCycleGAN-VC/VC2 to mel-spectrogram conversion. Through initial experiments, wediscovered that their direct applications compromised the time-frequencystructure that should be preserved during conversion. To remedy this, wepropose CycleGAN-VC3, an improvement of CycleGAN-VC2 that incorporatestime-frequency adaptive normalization (TFAN). Using TFAN, we can adjust thescale and bias of the converted features while reflecting the time-frequencystructure of the source mel-spectrogram. We evaluated CycleGAN-VC3 oninter-gender and intra-gender non-parallel VC. A subjective evaluation ofnaturalness and similarity showed that for every VC pair, CycleGAN-VC3outperforms or is competitive with the two types of CycleGAN-VC2, one of whichwas applied to mel-cepstrum and the other to mel-spectrogram. Audio samples areavailable athttp://www.kecl.ntt.co.jp/people/kaneko.takuhiro/projects/cyclegan-vc3/index.html.

CycleGAN-VC3: Examining and Improving CycleGAN-VCs for Mel-spectrogram Conversion