Sparse Gaussian Process Audio Source Separation Using Spectrum Priors in the Time-Domain

Gaussian process (GP) audio source separation is a time- domain approach that circumvents the inherent phase approx- imation issue of spectrogram based methods. Furthermore, through its kernel, GPs elegantly incorporate prior knowl- edge about the sources into the separation model. Despite these compelling advantages, the computational complexity of GP inference scales cubically with the number of audio samples. As a result, source separation GP models have been restricted to the analysis of short audio frames. We intro- duce an efficient application of GPs to time-domain audio source separation, without compromising performance. For this purpose, we used GP regression, together with spectral mixture kernels, and variational sparse GPs. We compared our method with LD-PSDTF (positive semi-definite tensor factorization), KL-NMF (Kullback-Leibler non-negative ma- trix factorization), and IS-NMF (Itakura-Saito NMF). Results show that the proposed method outperforms these techniques.