Search

Showing total 3 results
Start Over Topic embedding
3 results

1. Unsupervised regularization of the embedding extractor for robust language identification

Author

Raphaël Duroselle, Irina Illina, Denis Jouvet, Speech Modeling for Facilitating Oral-Based Communication (MULTISPEECH), Inria Nancy - Grand Est, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), Experiments presented in this paper were carried out using the Grid’5000 testbed, supported by a scientific interest group hosted by Inria and including CNRS, RENATER and several Universities as well as other organizations (see https://www.grid5000.fr). This work has been partly funded by the French Direction Générale de l'Armement., and Grid'5000

Subjects
Language identification, business.industry, Computer science, Pattern recognition, [INFO.INFO-NE]Computer Science [cs]/Neural and Evolutionary Computing [cs.NE], Regularization (mathematics), Extractor, Maximum mean discrepancy, Embedding, Labeled data, Artificial intelligence, business, Classifier (UML), Test data
Abstract

International audience; State-of-the-art spoken language identification systems are constituted of three modules: a frame-level feature extractor, a segment-level embedding extractor and a final classifier. The performance of these systems degrades when facing mismatch between training and testing data. Most domain adaptation methods focus on adaptation of the final classifier. In this article , we propose a model-based unsupervised domain adaptation of the segment-level embedding extractor. The approach consists in a modification of the loss function used for training the embedding extractor. We introduce a regularization term based on the maximum mean discrepancy loss. Experiments were performed on the RATS corpus with transmission channel mismatch between telephone and radio channels. We obtained the same language identification performance as supervised training on the target domains but without using labeled data from these domains.

Published
2020

2. Metric learning loss functions to reduce domain mismatch in the x-vector space for language recognition

Author

Raphaël Duroselle, Denis Jouvet, Irina Illina, Speech Modeling for Facilitating Oral-Based Communication (MULTISPEECH), Inria Nancy - Grand Est, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), Experiments presented in this paper were carried out using the Grid’5000 testbed, supported by a scientific interest group hosted by Inria and including CNRS, RENATER and several Universities as well as other organizations (see https://www.grid5000.fr). This work has been partly funded bythe French Direction Générale de l'Armement., and Grid'5000

Subjects
Computer science, domain adaptation, Speech recognition, metric learning, 020206 networking & telecommunications, 02 engineering and technology, 01 natural sciences, domain mismatch, language recognition, embedding, x-vector, Robustness (computer science), 0103 physical sciences, 0202 electrical engineering, electronic engineering, information engineering, Maximum mean discrepancy, Embedding, [INFO]Computer Science [cs], 010301 acoustics, Language recognition, Vector space
Abstract

International audience; State-of-the-art language recognition systems are based on dis-criminative embeddings called x-vectors. Channel and gender distortions produce mismatch in such x-vector space where em-beddings corresponding to the same language are not grouped in an unique cluster. To control this mismatch, we propose to train the x-vector DNN with metric learning objective functions. Combining a classification loss with the metric learning n-pair loss allows to improve the language recognition performance. Such a system achieves a robustness comparable to a system trained with a domain adaptation loss function but without using the domain information. We also analyze the mismatch due to channel and gender, in comparison to language proximity, in the x-vector space. This is achieved using the Maximum Mean Discrepancy divergence measure between groups of x-vectors. Our analysis shows that using the metric learning loss function reduces gender and channel mismatch in the x-vector space, even for languages only observed on one channel in the train set.

Published
2020

3. Transfer learning of the expressivity using flow metric learning in multispeaker text-to-speech synthesis

Author

Ajinkya Kulkarni, Vincent Colotte, Denis Jouvet, Jouvet, Denis, Speech Modeling for Facilitating Oral-Based Communication (MULTISPEECH), Inria Nancy - Grand Est, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Natural Language Processing & Knowledge Discovery (LORIA - NLPKD), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), Experiments presented in this paper were carried out using the Grid’5000 testbed, supported by a scientific interest group hosted by Inria and including CNRS, RENATER and several Universities as well as other organizations (see https://www.grid5000.fr)., Grid'5000, Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), and Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)

Subjects
Computer science, [INFO.INFO-TS] Computer Science [cs]/Signal and Image Processing, Mean opinion score, Speech recognition, deep metric learning, Inference, Speech synthesis, 02 engineering and technology, Latent variable, 010501 environmental sciences, [INFO] Computer Science [cs], computer.software_genre, expressivity, 01 natural sciences, [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, 0202 electrical engineering, electronic engineering, information engineering, [INFO]Computer Science [cs], variational autoencoder, text-to-speech, 0105 earth and related environmental sciences, Acoustic model, Speaker recognition, inverse autoregressive flow, Autoencoder, Metric (mathematics), Embedding, 020201 artificial intelligence & image processing, Transfer of learning, computer
Abstract

International audience; In this paper, we present a novel deep metric learning architecture along with variational inference incorporated in a paramet-ric multispeaker expressive text-to-speech (TTS) system. We proposed inverse autoregressive flow (IAF) as a way to perform the variational inference, thus providing flexible approximate posterior distribution. The proposed approach condition the text-to-speech system on speaker embeddings so that latent space represents the emotion as semantic characteristics. For representing the speaker, we extracted speaker em-beddings from the x-vector based speaker recognition model trained on speech data from many speakers. To predict the vocoder features, we used the acoustic model conditioned on the textual features as well as on the speaker embedding. We transferred the expressivity by using the mean of the latent variables for each emotion to generate expressive speech in different speaker's voices for which no expressive speech data is available. We compared the results obtained using flow-based variational inference with variational autoencoder as a base-line model. The performance measured by mean opinion score (MOS), speaker MOS, and expressive MOS shows that N-pair loss based deep metric learning along with IAF model improves the transfer of expressivity in the desired speaker's voice in synthesized speech.

Published
2020