Audio source localization and separation software
FASST: C++ toolbox for flexible audio source separation
Distributed under the QPL – Authors: Y Salaün, E Vincent
Reference
BSS Locate: Matlab toolbox for source localization
Distributed under the GPL – Authors: C Blandin, E Vincent, A Ozerov
Reference
SiSEC Reference Software: baseline software for source counting, mixing system estimation and source separation of under-determined mixtures
Distributed under the GPL – Authors: E Vincent, P Bofill, anonymous
Reference
Online FASST: online implementation of FASST
Distributed under the GPL – Authors: LSR Simon, E Vincent
Reference
Under-determined convolutive audio source separation by wideband convex optimization
Distributed under the CeCILL – Authors: M Kowalski, E Vincent, R Gribonval
Reference
Under-determined audio source separation by local Gaussian modeling for instantaneous mixtures and for convolutive mixtures
Distributed under the GPL – Authors: E Vincent, NQK Duong
Reference for instantaneous mixtures Reference for convolutive mixtures
Consistent Wiener filtering for audio source separation
Distributed under a proprietary research-only license – Authors: J Le Roux, E Vincent
Reference
Automatic speech recognition software
Kaldi to Matab conversion tools
Distributed under the GPL – Authors: E Vincent, S Watanabe
Uncertainty decoding for Kaldi
Distributed under the Apache license – Authors: Y Salaün, E Vincent
Full uncertainty decoding for HTK
Distributed under the BSD license – Authors: Y Salaün, E Vincent
Reference
Evaluation software
BSS Eval: Matlab toolbox for the objective evaluation of source separation algorithms
Distributed under the GPL – Authors: E Vincent, C Févotte, R Gribonval
Reference for bss_eval_mix.m Reference for bss_eval_sources.m Reference for bss_eval_images.m
PEASS: Matlab toolbox for perceptually-motivated objective evaluation of source separation algorithms
Distributed under the GPL – Authors: V Emiya
Reference
BSS Oracle: Matlab toolbox for the computation of upper performance bounds of source separation algorithms
Distributed under the GPL – Authors: E Vincent, R Gribonval, MD Plumbley
Reference
MUSHRAM: Matlab interface for MUSHRA listening tests
Distributed under the GPL – Author: E Vincent
Reference
Roomsimove: Matlab toolbox for the computation of simulated room impulse reponses for moving sources (this toolbox is also suitable for large-scale experiments with static sources)
Distributed under the GPL – Authors: E Vincent, DR Campbell
Other software
Music object-based coding and decoding: software for very low bit-rate coding of musical audio based on Bayesian harmonic models
Distributed under the GPL – Authors: E Vincent, MEP Davies
Reference
Melody estimation using loudness-weighted YIN or Bayesian harmonic models
Distributed under the GPL – Author: E Vincent
Reference
Multiple pitch estimation and note tracking using NMF under harmonicity and spectral smoothness constraints
Distributed under the GPL – Author: E Vincent
Reference
Quadratic ERB-scale time-frequency transform and multichannel filtering in the ERB-scale time-frequency domain
Distributed under the GPL – Author: E Vincent
Reference (Section II.A)
Data
Stereo Audio Source Separation Database part 1 and part 2: Instantaneous mixtures, synthetic convolutive mixtures and live recordings of speech or music sources for the development and testing of stereo source separation algorithms (see SiSEC wiki for a description of the data)
Distributed under the Creative Commons Attribution-NonCommercial-ShareAlike 2.0 License – Authors: E Vincent, H Sawada, S Araki, Another Dreamer, Alex Q
Reference
QUASI Music Source Separation Database: 12 multitrack songs, each mixed in several ways by a professional sound engineer
Distributed under various Creative Commons licenses – Authors: M Desnoues, J-L Durrieu, T Fillon, G Richard, O Le Blouch, E Vincent
Reference
DEMAND: a collection of multi-channel recordings of acoustic noise in diverse environments
Distributed under the Creative Commons Attribution-ShareAlike 3.0 Unported License – Authors: J Thiemann, N Ito, E Vincent
Reference
