Meta, the social networking giant formerly known as Facebook, has launched an open-source AI tool called AudioCraft. This tool is designed to generate realistic music and audio based on text prompts and audio signals. The announcement was made by Meta’s CEO, Mark Zuckerberg, on his Facebook page.
AudioCraft consists of three models: MusicGen, AudioGen, and EnCodec. MusicGen is trained with company-owned and licensed music to generate music from text prompts. AudioGen, on the other hand, is trained on public sound effects and generates audio from text prompts.
The company has also released an improved version of its EnCodec decoder, which allows higher quality music generation with fewer artifacts. Additionally, they have shared pre-trained AudioGen models that can generate environmental sounds and sound effects such as a dog barking, cars honking, or footsteps on a wooden floor.
Meta’s AudioCraft AI models are capable of producing high-quality audio with long-term consistency. They are designed to be user-friendly and easy to use.
According to Meta, AudioCraft simplifies the overall design of generative models for audio compared to previous work in the field. The company aims to empower users to push the limits and develop their own models using the provided code and models.
In an effort to advance the field of AI-generated audio and music, Meta has open-sourced these models, giving researchers and practitioners access to train their own models with their own datasets.
AudioCraft is a versatile tool that can be used for music, sound, compression, and generation all in one place. It provides a single code base for building better sound generators, compression algorithms, and music generators.
With the release of AudioCraft and the open-sourcing of its models, Meta hopes to contribute to the development and advancement of AI-generated audio and music. By allowing others to train their models, the company aims to foster innovation and collaboration in this field.
In conclusion, Meta’s launch of the open-source AI tool AudioCraft is set to revolutionize the generation of realistic music and audio. With its three models and user-friendly interface, AudioCraft provides a platform for both experienced researchers and newcomers to explore and create in the realm of AI-generated audio and music.
