The PEASS Software provides a set of perceptually motivated objective measures for the evaluation of audio source separation.
Similarly to BSS Eval, the distortion signal is decomposed into three components: target distortion, interference, artifacts. These components are then used to compute four quality scores, namely OPS (Overall Perceptual Score), TPS (Target-related Perceptual Score), IPS (Interference-related Perceptual Score), APS (Artifact-related Perceptual Score). These scores better correlate with human assessments than the SDR/ISR/SIR/SAR measures of BSS Eval.
This software is distributed under the terms of the GNU Public License version 3.
This is joint work by Valentin Emiya (INRIA, France), Emmanuel Vincent (INRIA, France), Niklas Harlander (University of Oldenburg, Germany), Volker Hohmann (University of Oldenburg, Germany). The reduction of the computation time of the decomposition step was proposed by Audionamix within the i3DMusic project.
When using this software, the following papers must be referred to:
 Valentin Emiya, Emmanuel Vincent, Niklas Harlander and Volker Hohmann, Subjective and objective quality assessment of audio source separation, IEEE Transactions on Audio, Speech and Language Processing, 19(7):2046-2057, 2011.
 Emmanuel Vincent, Improved perceptual metrics for the evaluation of audio source separation, 10th Int. Conf. on Latent Variable Analysis and Signal Separation (LVA/ICA 2012), submitted.
Back to the PEASS Toolkit