AudioCraft (Meta/Facebook)
AudioCraft is a library for audio processing and generation with deep learning. It was developed by Facebook AI Research and is open-source. AudioCraft features the state-of-the-art EnCodec audio compressor / tokenizer, along with MusicGen, a simple and controllable music generation LM with textual and melodic conditioning.
AudioCraft can be used to generate high-quality audio and music from text. It can also be used to compress audio files, making them smaller and easier to store and share. AudioCraft is still under development, but it has the potential to revolutionize the way we create and consume audio.
Here are some of the features of AudioCraft:
EnCodec: EnCodec is a neural audio codec that encodes audio signals into a sequence of discrete tokens. This makes it possible to store and transmit audio files more efficiently.
MusicGen: MusicGen is a music generation model that can be used to create new music from text. It can be conditioned on a variety of factors, such as the genre of music, the mood of the music, and the lyrics of the song.
AudioGen: AudioGen is a sound generation model that can be used to create new sounds from text. It can be conditioned on a variety of factors, such as the type of sound, the pitch of the sound, and the duration of the sound.
AudioCraft is a powerful tool for audio processing and generation. It is still under development, but it has the potential to revolutionize the way we create and consume audio.
Here are some links to learn more about AudioCraft:
AudioCraft website: https://audiocraft.metademolab.com/
AudioCraft GitHub repository: https://github.com/facebookresearch/audiocraft
AudioCraft paper: https://arxiv.org/abs/2206.07573
Last updated