Audiobox

Audiobox is a free and open-source AI voice and sound generation model launched by Meta on November 30, 2023. Its online web version was launched on December 11, allowing users to freely experience the capabilities of this model. Audiobox is the latest generation of audio generation model introduced by Meta following Voicebox. It can combine voice input and natural language text prompts to generate voices and sound effects, thus enabling the easy creation of realistic custom audio for various use cases.

The main functions of Audiobox.

1. Clone User Voice: Generate speech in the voice style of a user or in the style of any audio sample by recording voice.
2. Generate Human Voice from Text Description: Generate human voice by using text descriptions of the characteristics of voice styles and the acoustic environment.
3. Change Voice Style: Combine voice and text descriptions to change the existing voice style.
4. Generate Sound Effects from Text Description: Generate sound effects according to the text description of input voice characteristics.
5. Noise Cancellation: Provide the Magic Eraser function to eliminate transient noise in recordings.
6. Sound Filling: Replace a part of the audio with new sound according to text descriptions.
7. Audio Story Maker: Combine the above functions and use Audiobox Maker to create original and interesting audio stories.