Sound source modeling and separation in stereophonic mixtures: a PAN-driven approach generalizing ADRes

Abstract :

In this paper we will discuss a model aimed at improving the spectral data representation of stereophonic music signals (especially studio productions). In this work the stereo pair is thought as a single spectrum with additional dimensions, expressing the Interaural Intensity Difference (IID) data of each Fast Fourier Transform (FFT) bin. This view offers the ability to manipulate parts of the stereophonic spectrum in such ways that, under optimal circumstances, can result in a source separation algorithm.

Complete list of metadatas

https://hal.telecom-paristech.fr/hal-02286761
Contributor : Telecomparis Hal <>
Submitted on : Friday, September 13, 2019 - 4:09:20 PM
Last modification on : Thursday, October 17, 2019 - 12:37:02 PM

Identifiers

  • HAL Id : hal-02286761, version 1

Collections

Citation

Giorgio Presti, Davide Andrea Mauro. Sound source modeling and separation in stereophonic mixtures: a PAN-driven approach generalizing ADRes. IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), May 2014, Firenze, Italy. ⟨hal-02286761⟩

Share

Metrics

Record views

4