Audiobox is Meta’s new AI that debuts the ability to create conversations with voices from scratch
Meta has a new AI that is capable of cloning your voice in a matter of seconds. Audiobox is the name of this tool that uses AI to create voices from scratch, being able to imitate any type of voice with hardly any problems. This synthetic voice generator is available in Spain, and can be used by anyone.
Audiobox is the result of years of research and development by engineers at Meta, the company formerly known as Facebook. According to the official website, Audiobox is the first model that can generalize speech generation tasks without having been specifically trained for it and giving results never seen befores.
Meta launches Audiobox in Spain: this is its new AI tool with which to clone voices
Audiobox is based on an artificial intelligence model called Voicebox, which Meta developed and announced last June. The model uses a transformer architecture, similar to OpenAI’s ChatGPT or Google’s Bard, which allows it to learn in context and generalize across different text-to-speech tasks.
The operation is very simple, although at first it may seem overwhelming. What you have to keep in mind is that Meta itself has prepared a tutorial with everything you need to start using this tool and that it is highly recommended to follow.
What the tutorial explains is that You can create voices from scratch, requiring you to record yourself or another person reciting an example text that will appear on the screen. If you do not want to do this, you can always use the default predefined voices.
Having chosen a voice, Audiobox shows itself as an AI audio track editor. You can write a phrase for the voice to say and use different voices to form a conversation that, depending on what is written, may or may not make sense.
Besides, It has a section where you can add sound effects that will accompany the voices in order to create a conversation with greater atmosphere. Come on, Audiobox has the utility of a kind of audiobook in which each user can write their story.
The tool still needs a lot of development, in fact, the voices sound more robotic than they should and, apart from that, its launch has been made with informative intentions. Meta is clear that this tool will not be open source so you have to be cautious with the data offered.
To always be up to date with the latest in technology, subscribe to our official and verified Andro4all channel on WhatsApp.