Recently, Sun Yanzi's voice has been used for various covers. I also tried it and covered a small section of "Under the Fuji Mountain". This is for entertainment and learning purposes only.
Tools#
I referred to some tutorials from experts and used well-trained models. I mainly used these tools:
URV5: To cut the song to be replaced
Sun's voice model: Can be trained by oneself, but I directly used a well-trained model here
so-vits-svc: Voice training and conversion (web UI version by Bilibili user "羽毛布団")
Some thoughts#
Copyright issues are still unavoidable in the field of AI. Generally, vocal tones are not protected by copyright because it is difficult to give a unique identifier to a vocal tone. Just like this AI cover of a song, it can only be said to sound very similar to Sun's voice. As long as it is not acknowledged, it cannot be defined as using Sun's vocal tone. It is even possible to actively adjust the AI's vocal tone to achieve 90% similarity, making it completely indistinguishable.
Copyright protection applies to works, such as songs and articles, which are content creations. This is something to consider for commercialized AI voices. Either buy the rights or create your own.
AI voices have a wide range of applications and are suitable for creating audio content. One approach is to use existing well-known voices and blend them together. For example, using Sun's vocal tone but pronouncing certain words in the style of Jay Chou. In terms of content, text can be generated with the assistance of AIGC tools. Finally, it's important to consider the application scenarios and creativity.
I already have some preliminary ideas and will give them a try.