Výsledky vyhledávání - "RJ Skerry-Ryan"

Autor: Daisy Stanton, Matt Shannon, Soroosh Mariooryad, RJ Skerry-Ryan, Eric Battenberg, Tom Bagby, David Kao

This work explores the task of synthesizing speech in nonexistent human-sounding voices. We call this task "speaker generation", and present TacoSpawn, a system that performs competitively at this task. TacoSpawn is a recurrent attention-based text-t

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::f6e83b22f13a9e01602b62b69bb268cc
http://arxiv.org/abs/2111.05095

Zobrazit plný text záznamu

Parallel Tacotron 2: A Non-Autoregressive Neural TTS Model with Differentiable Duration Modeling

Autor: Yonghui Wu, RJ Skerry-Ryan, Jonathan Shen, Ye Jia, Yu Zhang, Isaac Elias, Heiga Zen

Publikováno v: Interspeech 2021.

This paper introduces Parallel Tacotron 2, a non-autoregressive neural text-to-speech model with a fully differentiable duration model which does not require supervised duration signals. The duration model is based on a novel attention mechanism and

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::8a573252a95a0f479450d28e3a7fc3ff
https://doi.org/10.21437/interspeech.2021-1461

Zobrazit plný text záznamu

Wave-Tacotron: Spectrogram-free end-to-end text-to-speech synthesis

Autor: Eric Battenberg, Soroosh Mariooryad, Ron Weiss, RJ Skerry-Ryan, Diederik P. Kingma

Publikováno v: ICASSP

We describe a sequence-to-sequence neural network which directly generates speech waveforms from text inputs. The architecture extends the Tacotron model by incorporating a normalizing flow into the autoregressive decoder loop. Output waveforms are m

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::9cdc85d56435b697621120f15b6834b4

Zobrazit plný text záznamu

Learning to Speak Fluently in a Foreign Language: Multilingual Speech Synthesis and Cross-Language Voice Cloning

Autor: Ron Weiss, Andrew Rosenberg, RJ Skerry-Ryan, Yu Zhang, Heiga Zen, Bhuvana Ramabhadran, Zhifeng Chen, Yonghui Wu, Ye Jia

Publikováno v: INTERSPEECH

We present a multispeaker, multilingual text-to-speech (TTS) synthesis model based on Tacotron that is able to produce high quality speech in multiple languages. Moreover, the model is able to transfer voices across languages, e.g. synthesize fluent

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::8526be8b339a406a410e03a2ae3a3a2e
http://arxiv.org/abs/1907.04448

Zobrazit plný text záznamu

Semi-supervised Training for Improving Data Efficiency in End-to-end Speech Synthesis

Autor: Yu Zhang, Yuxuan Wang, RJ Skerry-Ryan, Wei-Ning Hsu, Yu-An Chung

Publikováno v: ICASSP

Although end-to-end text-to-speech (TTS) models such as Tacotron have shown excellent results, they typically require a sizable set of high-quality pairs for training, which are expensive to collect. In this paper, we propose a semi-supervised traini

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::b26c52b21c3431da5d035a0ed2c1b372
https://doi.org/10.1109/icassp.2019.8683862

Zobrazit plný text záznamu

Location-Relative Attention Mechanisms For Robust Long-Form Speech Synthesis

Autor: Eric Battenberg, David T. H. Kao, Tom Bagby, Soroosh Mariooryad, RJ Skerry-Ryan, Daisy Stanton, Matt Shannon

Publikováno v: ICASSP

Despite the ability to produce human-level speech for in-domain text, attention-based end-to-end text-to-speech (TTS) systems suffer from text alignment failures that increase in frequency for out-of-domain text. We show that these failures can be ad

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::44e108a9536d4f771709c9eb9d0e9391

Zobrazit plný text záznamu

Complex Evolution Recurrent Neural Networks (ceRNNs)

Autor: Izhak Shafran, Tom Bagby, RJ Skerry-Ryan

Publikováno v: ICASSP

Unitary Evolution Recurrent Neural Networks (uRNNs) have three attractive properties: (a) the unitary property, (b) the complex-valued nature, and (c) their efficient linear operators [1]. The literature so far does not address - how critical is the

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::ceabbb082060f4b2e734e5e3baf06f27
https://doi.org/10.1109/icassp.2018.8462556

Zobrazit plný text záznamu

Predicting Expressive Speaking Style From Text In End-To-End Speech Synthesis

Autor: Daisy Stanton, Yuxuan Wang, RJ Skerry-Ryan

Publikováno v: SLT

Global Style Tokens (GSTs) are a recently-proposed method to learn latent disentangled representations of high-dimensional data. GSTs can be used within Tacotron, a state-of-the-art end-to-end text-to-speech synthesis system, to uncover expressive fa

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::4e7991f19ae638161f4ce6e5f4f733b6

Zobrazit plný text záznamu

Tacotron: Towards End-to-End Speech Synthesis

Autor: Rif A. Saurous, RJ Skerry-Ryan, Quoc V. Le, Yonghui Wu, Navdeep Jaitly, Daisy Stanton, Ron Weiss, Robert A. J. Clark, Yuxuan Wang, Yannis Agiomyrgiannakis, Ying Xiao, Zongheng Yang, Zhifeng Chen, Samy Bengio

Publikováno v: INTERSPEECH

A text-to-speech synthesis system typically consists of multiple stages, such as a text analysis frontend, an acoustic model and an audio synthesis module. Building these components often requires extensive domain expertise and may contain brittle de

Externí odkaz: https://explore.openaire.eu/search/publication?articleId=doi_dedup___::99c06db8493cf8f2b3b80aa98695161f
https://doi.org/10.21437/interspeech.2017-1452

Zobrazit plný text záznamu

Vyhledávací nástroje:

Upřesnit hledání