Kernel Spectrogram models for source separation

Abstract : In this study, we introduce a new framework called Kernel Additive Modelling for audio spectrograms that can be used for multichannel source separation. It assumes that the spectrogram of a source at any time-frequency bin is close to its value in a neighbourhood indicated by a source-specific proximity kernel. The rationale for this model is to easily account for features like periodicity, stability over time or frequency, self-similarity, etc. In many cases, such local dynamics are indeed much more natural to assess than any global model such as a tensor factorization. This framework permits one to use different proximity kernels for different sources and to estimate them blindly using their mixtures only. Estimation is performed using a variant of the kernel backfitting algorithm that allows for multichannel mixtures and permits parallelization. Experimental results on the separation of vocals from musical backgrounds demonstrate the efficiency of the approach.
Complete list of metadatas

Cited literature [28 references]  Display  Hide  Download

https://hal.inria.fr/hal-00959384
Contributor : Antoine Liutkus <>
Submitted on : Monday, February 16, 2015 - 12:28:14 AM
Last modification on : Thursday, July 4, 2019 - 11:00:07 AM
Long-term archiving on : Sunday, April 16, 2017 - 8:35:29 AM

File

KAM_HSCMAv2.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-00959384, version 4

Citation

Antoine Liutkus, Zafar Rafii, Bryan Pardo, Derry Fitzgerald, Laurent Daudet. Kernel Spectrogram models for source separation. HSCMA, May 2014, Nancy, France. ⟨hal-00959384v4⟩

Share

Metrics

Record views

720

Files downloads

671