The AI tool released by Deezer named Spleeter can split a track into four stems (stemming means to isolate vocal tracks from a mixed-down audio file) 100 times faster than real time claims the company.
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.
At present, the software lacks a graphical user interface. As the software is open source, a streamlined user interface can be developed. Though many techniques exist to isolate vocal tracks from a mixed-down audio file (called ‘stems’), most of them are difficult, time-consuming and of low quality.
Python and TensorFlow
Spleeter is an audio separation library built on Python and the machine learning platform TensorFlow that can split a track at many times the speed of the original audio says the company. Spleeter works on ‘source separation’ by using machine-learning to separate tracks into their composite stems like bass, drums, vocals, guitar and piano.