Deezer, the French online music streaming service has announced that it is releasing Spleeter – an open-source library for sound source separation.
Sound source separation is an important task in signal processing and it has a large number of applications, for example in remixes, mixing, active listening, transcription, etc. A large number of methods have been proposed in the past but still, sound separation remains a challenging task.
Deezer Open Source Download
If you want to debug or build it from source: there is a docker-compose file in the docker directory. The docker/downloads directory is mounted into the container and will be used as download directory. You have to check the permissions of the docker/downloads directory as docker mounts it with the same owner/group/permissions as on the host.
According to Deezer’s blog post, their sound separation model Spleeter performs at least as good as the best proposed algorithms currently available. They decided to open-source the model together with a library also called Spleeter.
- French streaming platform Deezer has launched an AI tool named Spleeter that can isolate vocal and instrumental tracks quickly and separate a song into two, four or five separate audio tracks. It has released the software as an open-source package on GitHub though it was originally developed for research purposes.
- The usage is quite similar to JCenter or MavenCentral, except that Nexus (open source version) is installed on Deezer servers and is reachable only within Deezer. The publishing process is handled with our Continuous Integration server (Jenkins in our case). A dedicated job takes care of building the library and uploading it to the Nexus.
- Deezer, the French online music streaming service has announced that it is releasing Spleeter – an open-source library for sound source separation. Sound source separation is an important task in signal processing and it has a large number of applications, for example in remixes, mixing, active listening, transcription, etc.
- Download Deezer on Windows, Mac, iOS, Android, and all your devices, and listen to over 73 million songs in streaming and offline.
The library is written in Python and built on top of Tensorflow. It allows for easy training of source separation models and it contains and already pre-trained state-of-the-art sound separation model from Deezer. The library can work within a GPU accelerated environment and achieve 100x faster than real-time processing for sound source separation. Therefore, Spleeter can also be used to process large datasets.
Several different models based were included in the Spleeter library: “vocals (singing voice)/accompaniment separation (2 source), “vocals/drums/bass/other” separation (4-source) and “vocals/ drums/bass/piano/other”, 5-source separation. The 2-source and 4-source models achieve state-of-the-art performance on the musdb dataset.
More about Spleeter can be read in the official blog post or in the library’s documentation.
The models available are:
- Vocals (singing voice) / accompaniment separation (2 stems)
- Vocals / drums / bass / other separation (4 stems)
- Vocals / drums / bass / piano / other separation (5 stems)
2 stems and 4 stems models have state of the art performances on the musdb dataset. Spleeter is also very fast as it can perform separation of audio files to 4 stems 100x faster than real-time when run on a GPU.
Deezer Open Source Music
We designed Spleeter so you can use it straight from command line as well as directly in your own development pipeline as a Python library. It can be installed with Conda, with pip or be used with Docker.