Filters








802 Hits in 8.1 sec

Multichannel audio source separation: Variational inference of time-frequency sources from time-domain observations

Simon Leglaive, Roland Badeau, Gael Richard
2017 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
A great number of methods for multichannel audio source separation are based on probabilistic approaches in which the sources are modeled as latent random variables in a Time-Frequency (TF) domain.  ...  The TF latent sources are then inferred from the TF mixture observations. In this paper we propose to infer the TF latent sources from the time-domain observations.  ...  Source separation is commonly achieved in a Time-Frequency (TF) domain because it provides a meaningful and often sparse representation of the source signals.  ... 
doi:10.1109/icassp.2017.7951791 dblp:conf/icassp/LeglaiveBR17 fatcat:rqg3m7a64fdr3l5n7tclqrcfeq

Alpha-stable multichannel audio source separation

Simon Leglaive, Umut Simsekli, Antoine Liutkus, Roland Badeau, Gael Richard
2017 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
In this paper, we focus on modeling multichannel audio signals in the short-time Fourier transform domain for the purpose of source separation.  ...  We develop a Monte Carlo Expectation-Maximization algorithm for inferring the sources from the proposed model.  ...  INTRODUCTION Multichannel audio source separation is the task that aims to recover a set of source audio signals from an observed mixture signal that has multiple channels (e.g. stereo audio).  ... 
doi:10.1109/icassp.2017.7952221 dblp:conf/icassp/LeglaiveSLBR17 fatcat:qxrxritdrjdlvk5ajwvs5zwrme

Student's t Source and Mixing Models for Multichannel Audio Source Separation

Simon Leglaive, Roland Badeau, Gael Richard
2018 IEEE/ACM Transactions on Audio Speech and Language Processing  
From this model, we develop a variational inference algorithm in order to perform source separation.  ...  Index Terms-Audio source separation, multichannel reverberant mixtures, Student's t distribution, statistical room acoustics, non-negative matrix factorization, variational inference.  ...  INTRODUCTION A UDIO source separation consists in recovering a set of audio source signals from the observation of a mixture signal.  ... 
doi:10.1109/taslp.2018.2813011 fatcat:gmiobaexnjawnbidhcp3koezmq

Semi-Blind Student'S T Source Separation For Multichannel Audio Convolutive Mixtures

Roland Badeau, Simon Leglaive, Gael Richard
2018 Zenodo  
Publication in the conference proceedings of EUSIPCO, Kos island, Greece, 2017  ...  INTRODUCTION Multichannel audio source separation aims to recover a set of audio source signals from several observed mixtures.  ...  We then use the time-domain observations to infer the TF latent source variables. The inference relies on a variational expectation-maximization (VEM) algorithm [20] .  ... 
doi:10.5281/zenodo.1159327 fatcat:pvhkv7gurvbx7gq5dcqchuj5zu

Gaussian Model Based Multichannel Separation [chapter]

Alexey Ozerov, Hirokazu Kameoka
2018 Audio Source Separation and Speech Enhancement  
must then be solved in order to align the separated components in different frequency bins that originate from the same source.  ...  the predominant source, i.e., the most active source in time-frequency bin (n, f ).  ...  Now, let us turn to the cost function (1.47) for multichannel NMF where (1.92) We can confirm that when the number of channels and sources is I = 1 and J = 1, respectively, and φ jk = 1, this cost function  ... 
doi:10.1002/9781119279860.ch14 fatcat:5cesv5jtjbhh3pog5faldytc7q

Semi-blind student's t source separation for multichannel audio convolutive mixtures

Simon Leglaive, Roland Badeau, Gael Richard
2017 2017 25th European Signal Processing Conference (EUSIPCO)  
Index Terms-Under-determined audio source separation, multichannel convolutive mixture, Student's t distribution, nonnegative matrix factorization, variational inference.  ...  This paper addresses the problem of multichannel audio source separation in under-determined convolutive mixtures. We target a semi-blind scenario assuming that the mixing filters are known.  ...  INTRODUCTION Multichannel audio source separation aims to recover a set of audio source signals from several observed mixtures.  ... 
doi:10.23919/eusipco.2017.8081612 dblp:conf/eusipco/LeglaiveBR17 fatcat:dklpqnujzja4jdchddsaz4opte

Joint audio source separation and dereverberation based on multichannel factorial hidden Markov model

Takuya Higuchi, Hirokazu Kameoka
2014 2014 IEEE International Workshop on Machine Learning for Signal Processing (MLSP)  
The multichannel NMF assumes that an observed signal is a mixture of a limited number of source signals each of which has a static power spectral density scaled by a time-varying amplitude.  ...  We have previously proposed an extension of this approach, in which the variations over time of the spectral density and the total power of each source is modeled by a hidden Markov model (HMM).  ...  Therefore we approximately express the observed signals as a form of a convolution of the frequency array response and the source signal in the time-frequency domain: y(ω k , t l ) ≈ I ∑ i=1 N ∑ n=0 a  ... 
doi:10.1109/mlsp.2014.6958927 dblp:conf/mlsp/HiguchiK14 fatcat:hnlxs2ctkvexre5szejio23jvm

Deep unfolding for multichannel source separation

Scott Wisdom, John Hershey, Jonathan Le Roux, Shinji Watanabe
2016 2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
We unfold a multichannel Gaussian mixture model (MCGMM), resulting in a deep MCGMM computational network that directly processes complex-valued frequency-domain multichannel audio and has an architecture  ...  Deep unfolding has recently been proposed to derive novel deep network architectures from model-based approaches. In this paper, we consider its application to multichannel source separation.  ...  [16] used a CNN-DNN for acoustic modeling on raw time-domain multichannel audio. Nugraha et al.  ... 
doi:10.1109/icassp.2016.7471649 dblp:conf/icassp/WisdomHRW16 fatcat:ggmgp25qczcbfgz7vzh4qglbva

Audio source separation into the wild [chapter]

Laurent Girin, Sharon Gannot, Xiaofei Li
2019 Multimodal Behavior Analysis in the Wild  
This review chapter is dedicated to multichannel audio source separation in real-life environment. We explore some of the major achievements in the field and discuss some of the remaining challenges.  ...  We will explore several important practical scenarios, e.g. moving sources and/or microphones, varying number of sources and sensors, high reverberation levels, spatially diffuse sources, and synchronization  ...  Multichannel audio source separation In this section, we briefly present the fundamentals of multichannel audio source separation.  ... 
doi:10.1016/b978-0-12-814601-9.00022-5 fatcat:zvc74jczyvg67ifxq3seouigfi

Neural Full-Rank Spatial Covariance Analysis for Blind Source Separation

Yoshiaki Bando, Kouhei Sekiguchi, Yoshiki Masuyama, Aditya Arie Nugraha, Mathieu Fontaine, Kazuyoshi Yoshii
2021 IEEE Signal Processing Letters  
This paper describes a neural blind source separation (BSS) method based on amortized variational inference (AVI) of a non-linear generative model of mixture signals.  ...  Specifically, we introduce a neural mixture-to-feature inference model that directly infers the latent features from the observed mixture and integrate it with a neural feature-to-mixture generative model  ...  Its trained decoder is combined with a spatial model to separate source signals by estimating the latent feature vectors of sound sources from a multichannel mixture.  ... 
doi:10.1109/lsp.2021.3101699 fatcat:o37fcxqdg5f3phbnxhwbdvkhai

Sound Source Separation [chapter]

G. Evangelista, S. Marchand, M. D. Plumbley, E. Vincent
2011 DAFX: Digital Audio Effects  
Acknowledgements The authors would like to thank Beiming Wang, Harald Viste and Mathieu Lagrange for assistance with some of the figures in this chapter.  ...  Multichannel Separation of Overlapping Harmonics The separation of time-frequency overlapping source components is a very challenging problem.  ...  Chapter 1 Sound Source Separation For each bin of the discrete time-frequency spectrum, let us now consider the observed magnitude in this frequency bin as time goes by.  ... 
doi:10.1002/9781119991298.ch14 fatcat:afqzxgyhzre2znstqgtgdh3uqi

Adaptive processing and learning for audio source separation

Jen-Tzung Chien, Hiroshi Sawada, Shoji Makino
2013 2013 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference  
This paper overviews a series of recent advances in adaptive processing and learning for audio source separation.  ...  In real world, speech and audio signal mixtures are observed in reverberant environments. Sources are usually more than mixtures.  ...  Therefore, we need to align the permutation ambiguity of the ICA or GMM results in each frequency bin so that a separated signal in the time domain contains frequency components from the same source signal  ... 
doi:10.1109/apsipa.2013.6694302 dblp:conf/apsipa/ChienSM13 fatcat:ydoxdq474ng6bhlvavqdvh2twq

An em algorithm for audio source separation based on the convolutive transfer function

Xiaofei Li, Laurent Girin, Radu Horaud
2017 2017 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA)  
This paper addresses the problem of audio source separation from (possibly under-determined) multichannel convolutive mixtures.  ...  We propose a separation method based on the convolutive transfer function (CTF) in the short-time Fourier transform domain.  ...  INTRODUCTION In this paper we address the problem of multichannel audio source separation (MASS) from (possibly underdetermined) convolutive mixtures.  ... 
doi:10.1109/waspaa.2017.8169994 dblp:conf/waspaa/LiGH17 fatcat:r3hhqjlprzeoxc4eregbezry6a

Separating time-frequency sources from time-domain convolutive mixtures using non-negative matrix factorization

Simon Leglaive, Roland Badeau, Gael Richard
2017 2017 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA)  
Index Terms-Audio source separation, reverberant mixtures, non-negative matrix factorization, variational inference.  ...  Source separation is performed from the time-domain mixture signals in order to accurately model the convolutive mixing process.  ...  INTRODUCTION Multichannel audio source separation consists in recovering several source signals from the observation of a mixture recorded with multiple microphones.  ... 
doi:10.1109/waspaa.2017.8170036 dblp:conf/waspaa/LeglaiveBR17 fatcat:ljwtf46ptfeylmh6y4i4u2yjzy

Two-stage blind audio source counting and separation of stereo instantaneous mixtures using Bayesian tensor factorisation

Sayeh Mirzaei, Hugo Van Hamme, Yaser Norouzi
2015 IET Signal Processing  
The performance of the source separation is measured by obtaining the existing metrics for multichannel blind source separation evaluation.  ...  In this paper, the authors address the tasks of audio source counting and separation for two-channel instantaneous mixtures. This goal is achieved in two steps.  ...  This is manifested by observing the corresponding BSS evaluation metrics and also the separated signals in the time domain.  ... 
doi:10.1049/iet-spr.2014.0404 fatcat:tlvupluhwre5jh7acb2e5zs5me
« Previous Showing results 1 — 15 out of 802 results