Algorithms for audio inpainting based on probabilistic nonnegative matrix factorization

Autor: Ondřej Mokrý, Paul Magron, Thomas Oberlin, Cédric Févotte

Publikováno v: Signal Processing
Signal Processing, 2023

International audience; Audio inpainting, i.e., the task of restoring missing or occluded audio signal samples, usually relies on sparse representations or autoregressive modeling. In this paper, we propose to structure the spectrogram with nonnegati

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::abdd17cbe00e59cc0ffc3d447feea709
https://inria.hal.science/hal-03708613v2/document

Phase recovery with Bregman divergences for audio source separation

Autor: Pierre-Hugo Vial, Cédric Févotte, Paul Magron, Thomas Oberlin

Publikováno v: IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Jun 2021, Toronto, Canada
ICASSP

International audience; Time-frequency audio source separation is usually achieved by estimating the short-time Fourier transform (STFT) magnitude of each source, and then applying a phase recovery algorithm to retrieve time-domain signals. In partic

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::c205767866ab2d87d1884271095485b0
https://hal.science/hal-03049800

Online Spectrogram Inversion for Low-Latency Audio Source Separation

Autor: Paul Magron, Tuomas Virtanen

Publikováno v: IEEE Signal Processing Letters
IEEE Signal Processing Letters, Institute of Electrical and Electronics Engineers, 2020, 27, pp.306-310. ⟨10.1109/LSP.2020.2970310⟩

International audience; Audio source separation is usually achieved by estimating the short-time Fourier transform (STFT) magnitude of each source, and then applying a spectrogram inversion algorithm to retrieve time-domain signals. In particular, th

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::6187f4a2b86430709bbb3467e6a21cec
https://hal.inria.fr/hal-03132170/file/Online_MISI.pdf

Language Modelling for Sound Event Detection with Teacher Forcing and Scheduled Sampling

Autor: Paul Magron, Shayan Gharib, Tuomas Virtanen, Konstantinos Drossos

Publikováno v: IEEE AASP Challenge on Detection and Classification of Acoustic Scenes and Events Workshops (DCASE 2019)
IEEE AASP Challenge on Detection and Classification of Acoustic Scenes and Events Workshops (DCASE 2019), Oct 2019, New York, United States
Tampere University

A sound event detection (SED) method typically takes as an input a sequence of audio frames and predicts the activities of sound events in each frame. In real-life recordings, the sound events exhibit some temporal structure: for instance, a "car hor

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::710c9b3e7031920af26882172b20b4d8
http://arxiv.org/abs/1907.08506

Harmonic-Percussive Source Separation with Deep Neural Networks and Phase Recovery

Autor: Konstantinos Drossos, Tuomas Virtanen, Paul Magron, Stylianos Ioannis Mimilakis

Publikováno v: International Workshop on Acoustic Signal Enhancement
International Workshop on Acoustic Signal Enhancement, Sep 2018, Tokyo, Japan
IWAENC
2018 16th International Workshop on Acoustic Signal Enhancement (IWAENC)

International audience; Harmonic/percussive source separation (HPSS) consists in separating the pitched instruments from the percussive parts in a music mixture. In this paper, we propose to apply the recently introduced Masker-Denoiser with twin net

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::dfc4b662cd810553dbef4e2e8141c807
https://hal.archives-ouvertes.fr/hal-01812225v2/document