The PEASS Toolkit - Perceptual Evaluation methods for Audio Source Separation

Valentin Emiya 1 Emmanuel Vincent 1 Niklas Harlander 2 Volker Hohmann 2
1 METISS - Speech and sound data modeling and processing
IRISA - Institut de Recherche en Informatique et Systèmes Aléatoires, Inria Rennes – Bretagne Atlantique
Abstract : We present the PEASS Toolkit designed for the perceptual evaluation of audio source separation. It includes a MUSHRA-based listening-test protocol adapted for the subjective evaluation of audio source separation; a set of subjective measures - sounds and subjective scores - obtained by applying the proposed test protocol; and a set of auditory-motivated objective measures based on a new algorithm to decompose the estimation error into some distortion components (target distortion, interference, artefacts) and on integrating the outputs of the auditory-based objective measure PEMO-Q. This presentation will be the opportunity to detail the recent release of the PEASS toolkit, as well as its use in the Signal Separation Evaluation Campaign (SiSEC) in 2010.
Complete list of metadatas

https://hal.inria.fr/inria-00545477
Contributor : Valentin Emiya <>
Submitted on : Wednesday, December 7, 2011 - 2:03:14 PM
Last modification on : Thursday, March 21, 2019 - 2:20:42 PM
Long-term archiving on : Thursday, March 8, 2012 - 2:20:16 AM

File

2010-09-27_LVA_PEASS_emiya.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : inria-00545477, version 1

Citation

Valentin Emiya, Emmanuel Vincent, Niklas Harlander, Volker Hohmann. The PEASS Toolkit - Perceptual Evaluation methods for Audio Source Separation. 9th Int. Conf. on Latent Variable Analysis and Signal Separation, Sep 2010, Saint-Malo, France. 2010. ⟨inria-00545477⟩

Share

Metrics

Record views

498

Files downloads

794