Monaural Singing Voice Separation with Skip-Filtering Connections and Recurrent Inference of Time-Frequency Mask
This work addresses the problem of separating singing voices from monaural audio mixtures for audio processing applications, representing an incremental improvement over existing methods.
The paper tackles monaural singing voice separation by learning and optimizing a source-dependent mask during training, eliminating the need for post-processing steps like generalized Wiener filtering. The method achieved an increase of 0.49 dB in signal-to-distortion ratio and 0.30 dB in signal-to-interference ratio compared to previous state-of-the-art approaches.
Singing voice separation based on deep learning relies on the usage of time-frequency masking. In many cases the masking process is not a learnable function or is not encapsulated into the deep learning optimization. Consequently, most of the existing methods rely on a post processing step using the generalized Wiener filtering. This work proposes a method that learns and optimizes (during training) a source-dependent mask and does not need the aforementioned post processing step. We introduce a recurrent inference algorithm, a sparse transformation step to improve the mask generation process, and a learned denoising filter. Obtained results show an increase of 0.49 dB for the signal to distortion ratio and 0.30 dB for the signal to interference ratio, compared to previous state-of-the-art approaches for monaural singing voice separation.