Deep Karaoke: Extracting Vocals from Musical Mixtures Using a Convolutional Deep Neural Network
This addresses the challenge of source separation in music for applications like karaoke, but it is incremental as it adapts existing DNN methods from speech to music.
The paper tackled the problem of extracting singing vocals from musical mixtures by training a convolutional deep neural network to estimate ideal binary masks, achieving results that contrast with traditional linear methods for potential karaoke applications.
Identification and extraction of singing voice from within musical mixtures is a key challenge in source separation and machine audition. Recently, deep neural networks (DNN) have been used to estimate 'ideal' binary masks for carefully controlled cocktail party speech separation problems. However, it is not yet known whether these methods are capable of generalizing to the discrimination of voice and non-voice in the context of musical mixtures. Here, we trained a convolutional DNN (of around a billion parameters) to provide probabilistic estimates of the ideal binary mask for separation of vocal sounds from real-world musical mixtures. We contrast our DNN results with more traditional linear methods. Our approach may be useful for automatic removal of vocal sounds from musical mixtures for 'karaoke' type applications.