Zobrazeno 1 - 10
of 13
pro vyhledávání: '"Natalia Tomashenko"'
This paper presents a study on the use of federated learning to train an ASR model based on a wav2vec 2.0 model pre-trained by self supervision. Carried out on the well-known TED-LIUM 3 dataset, our experiments show that such a model can obtain, with
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::1e328d3a34dbbb67e0451b9427824bfe
Autor:
Brij Mohan Lal Srivastava, Mohamed Maouche, Md Sahidullah, Emmanuel Vincent, Aurelien Bellet, Marc Tommasi, Natalia Tomashenko, Xin Wang, Junichi Yamagishi
Publikováno v:
IEEE/ACM Transactions on Audio, Speech and Language Processing
IEEE/ACM Transactions on Audio, Speech and Language Processing, 2022, ⟨10.1109/TASLP.2022.3190741⟩
IEEE/ACM Transactions on Audio, Speech and Language Processing, 2022, ⟨10.1109/TASLP.2022.3190741⟩
International audience; We study the scenario where individuals (speakers) contribute to the publication of an anonymized speech corpus. Data users then leverage this public corpus to perform downstream tasks (such as training automatic speech recogn
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::078334922c49b17cd7378abbb51a9438
https://inria.hal.science/hal-03197376v3/document
https://inria.hal.science/hal-03197376v3/document
Publikováno v:
JEP 2022
JEP 2022, Jun 2022, île de Noirmoutier, France
JEP 2022, Jun 2022, île de Noirmoutier, France
National audience; Plusieurs services intégrés dans notre vie quotidienne utilisent la reconnaissance automatique de la parole (Apple-Siri, Amazon-Alexa...). Ces services s'appuient sur des modèles entraînés sur une grande quantité de données
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::488074dda5e21679943deabde2e9e10f
http://hdl.handle.net/20.500.12210/80092
http://hdl.handle.net/20.500.12210/80092
Publikováno v:
XXXIVe Journées d'Études sur la Parole -- JEP 2022.
Self-supervised models for speech processing emerged recently as popular foundation blocks in speech processing pipelines. These models are pre-trained on unlabeled audio data and then used in speech processing downstream tasks such as automatic spee
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::8401d011e50eb17e736d1c2b6939953c
http://arxiv.org/abs/2204.01397
http://arxiv.org/abs/2204.01397
In our previous work, we proposed a language-independent speaker anonymization system based on self-supervised learning models. Although the system can anonymize speech data of any language, the anonymization was imperfect, and the speech content of
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::775695c7874b640e4ac1f16ee1843d42
http://arxiv.org/abs/2203.14834
http://arxiv.org/abs/2203.14834
Publikováno v:
ICASSP 2022
ICASSP 2022, 2022, Singapour, Singapore
ICASSP 2022, 2022, Singapour, Singapore
This paper investigates methods to effectively retrieve speaker information from the personalized speaker adapted neural network acoustic models (AMs) in automatic speech recognition (ASR). This problem is especially important in the context of feder
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::0d307046ec338ddc49eb994ffd835b82
https://hal.science/hal-03539742v2/document
https://hal.science/hal-03539742v2/document
Autor:
Hang Le, Sina Alisamir, Marco Dinarelli, Fabien Ringeval, Solène Evain, Ha Nguyen, Marcely Zanon Boito, Salima Mdhaffar, Ziyi Tong, Natalia Tomashenko, Titouan Parcollet, Alexandre Allauzen, Yannick Estève, Benjamin Lecouteux, François Portet, Solange Rossato, Didier Schwab, Laurent Besacier
Publikováno v:
HAL
L'apprentissage autosupervisé a apporté des améliorations remarquables dans de nombreux domaines tels que la vision par ordinateur ou le traitement de la langue et de la parole, en exploitant de grandes quantités de données non étiquetées. Dan
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::5b45625171543c36fa01b1db28207288
https://hal.archives-ouvertes.fr/hal-03706952
https://hal.archives-ouvertes.fr/hal-03706952
Publikováno v:
IEEE ICASSP 2022
IEEE ICASSP 2022, 2022, Singapour, Singapore
IEEE ICASSP 2022, 2022, Singapour, Singapore
International audience; The widespread of powerful personal devices capable of collecting voice of their users has opened the opportunity to build speaker adapted speech recognition system (ASR) or to participate to collaborative learning of ASR. In
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::6daf7cfd935e69d12da1144313a84aa4
https://hal.science/hal-03539741/document
https://hal.science/hal-03539741/document
Autor:
Natalia Tomashenko, Xin Wang, Emmanuel Vincent, Jose Patino, Brij Mohan Lal Srivastava, Paul-Gauthier Noé, Andreas Nautsch, Nicholas Evans, Junichi Yamagishi, Benjamin O’Brien, Anaïs Chanclu, Jean-François Bonastre, Massimiliano Todisco, Mohamed Maouche
Publikováno v:
Computer Speech and Language
Computer Speech and Language, 2022, 74, pp.101362. ⟨10.1016/j.csl.2022.101362⟩
Computer Speech and Language, Elsevier, 2022, 74, pp.101362. ⟨10.1016/j.csl.2022.101362⟩
Computer Speech and Language, 2022, 74, pp.101362. ⟨10.1016/j.csl.2022.101362⟩
Computer Speech and Language, Elsevier, 2022, 74, pp.101362. ⟨10.1016/j.csl.2022.101362⟩
This paper presents the results and analyses stemming from the first VoicePrivacy 2020 Challenge which focuses on developing anonymization solutions for speech technology. We provide a systematic overview of the challenge design with an analysis of s
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::4956d684be2723627c42a0e95911eb4c
https://hal.archives-ouvertes.fr/hal-03332224
https://hal.archives-ouvertes.fr/hal-03332224