Audio AIGC Model
Currently, there are no good basic models available in the audio field.
| Basic Model | Release Date | Publisher | Note |
|---|---|---|---|
| dance-diffusion | 2022.09 | harmonai | |
| audio-diffusion | 2022.08 | teticio | |
| riffusion | 2022.12 | Seth Forsgren | Generates a spectrum using a diffusion model and converts it to music |
| audioldm | 2023.01 | haoheliu | |
| bark | 2023.04 | suno.ai |
Currently available open-source models have a certain distance from application. If you want to experience better effects, you can take a look at mubert.
In addition to models for generating music or sound from scratch, another technology that is currently approaching the application threshold is voice conversion.
Voice conversion is a technology that can modify the speech of the source speaker to sound like the voice of another target speaker.
Currently, a well-known product is so-vits-svc for Chinese.