admin管理员组

文章数量:1123097

I am trying to do accent conversion and I used mfccs as a feature. Now, I have converted mfccs and I want to check how it sounds. For that, I need to synthesize speech from the given mfccs. Do anyone have idea on how can I generate high quality audio from the mfccs? Is there any matlab code or python code for that? Or is there any tutorial on that?

I tried using griffin-lim algorithm which is not very effective. I also found Hifi-GAN which synthesizes great quality audio but it works with melspectrograms not with mfccs. I wanted to use a neural vocoder if there is any to generate speech from the mfccs.

本文标签： deep learningHow to synthesize high quality speech from mfccsStack Overflow

版权声明：本文标题：deep learning - How to synthesize high quality speech from mfccs? - Stack Overflow 内容由网友自发贡献，该文观点仅代表作者本人， 转载请联系作者并注明出处：http://www.betaflare.com/web/1736546946a1944459.html， 本站仅提供信息存储空间服务，不拥有所有权，不承担相关法律责任。如发现本站有涉嫌抄袭侵权/违法违规的内容，一经查实，本站将立刻删除。