Scipediacontent: Scipediacontent moved page Draft Content 245975100 to Wang et al 2019b

2021-02-01T19:47:03Z

Scipediacontent moved page Draft Content 245975100 to Wang et al 2019b

Scipediacontent: Created page with " == Abstract == End-to-end speech synthesis is a promising approach that directly converts raw text to speech. Although it was shown that Tacotron2 outperforms classical pipe..."

2021-02-01T19:46:58Z

Created page with " == Abstract == End-to-end speech synthesis is a promising approach that directly converts raw text to speech. Although it was shown that Tacotron2 outperforms classical pipe..."

New page

== Abstract ==

End-to-end speech synthesis is a promising approach that directly converts raw text to speech. Although it was shown that Tacotron2 outperforms classical pipeline systems with regards to naturalness in English, its applicability to other languages is still unknown. Japanese could be one of the most difficult languages for which to achieve end-to-end speech synthesis, largely due to its character diversity and pitch accents. Therefore, state-of-the-art systems are still based on a traditional pipeline framework that requires a separate text analyzer and duration model. Towards end-to-end Japanese speech synthesis, we extend Tacotron to systems with self-attention to capture long-term dependencies related to pitch accents and compare their audio quality with classical pipeline systems under various conditions to show their pros and cons. In a large-scale listening test, we investigated the impacts of the presence of accentual-type labels, the use of force or predicted alignments, and acoustic features used as local condition parameters of the Wavenet vocoder. Our results reveal that although the proposed systems still do not match the quality of a top-line pipeline system for Japanese, we show important stepping stones towards end-to-end Japanese speech synthesis.

Comment: to be appeared at ICASSP 2019

== Original document ==

The different versions of the original document can be found in:

* [http://arxiv.org/abs/1810.11960 http://arxiv.org/abs/1810.11960]

* [http://arxiv.org/pdf/1810.11960 http://arxiv.org/pdf/1810.11960]

* [http://xplorestaging.ieee.org/ielx7/8671773/8682151/08682353.pdf?arnumber=8682353 http://xplorestaging.ieee.org/ielx7/8671773/8682151/08682353.pdf?arnumber=8682353],
: [http://dx.doi.org/10.1109/icassp.2019.8682353 http://dx.doi.org/10.1109/icassp.2019.8682353]

* [https://dblp.uni-trier.de/db/conf/icassp/icassp2019.html#YasudaWTY19 https://dblp.uni-trier.de/db/conf/icassp/icassp2019.html#YasudaWTY19],
: [https://arxiv.org/abs/1810.11960 https://arxiv.org/abs/1810.11960],
: [https://arxiv.org/pdf/1810.11960.pdf https://arxiv.org/pdf/1810.11960.pdf],
: [https://ieeexplore.ieee.org/document/8682353 https://ieeexplore.ieee.org/document/8682353],
: [https://academic.microsoft.com/#/detail/2963945466 https://academic.microsoft.com/#/detail/2963945466]

* [https://www.research.ed.ac.uk/portal/en/publications/investigation-of-enhanced-tacotron-texttospeech-synthesis-systems-with-selfattention-for-pitch-accent-language(a0c1b235-3613-4ea8-a511-cb1692d1fd1b).html https://www.research.ed.ac.uk/portal/en/publications/investigation-of-enhanced-tacotron-texttospeech-synthesis-systems-with-selfattention-for-pitch-accent-language(a0c1b235-3613-4ea8-a511-cb1692d1fd1b).html],
: [https://doi.org/10.1109/ICASSP.2019.8682353 https://doi.org/10.1109/ICASSP.2019.8682353],
: [http://hdl.handle.net/20.500.11820/a0c1b235-3613-4ea8-a511-cb1692d1fd1b http://hdl.handle.net/20.500.11820/a0c1b235-3613-4ea8-a511-cb1692d1fd1b],
: [https://www.pure.ed.ac.uk/ws/files/99996986/Investigation_of_enhanced_Tacotron_YASUDA_DoA020219_AFV.pdf https://www.pure.ed.ac.uk/ws/files/99996986/Investigation_of_enhanced_Tacotron_YASUDA_DoA020219_AFV.pdf],
: [https://ieeexplore.ieee.org/document/8682353 https://ieeexplore.ieee.org/document/8682353]

← Older revision	Revision as of 19:47, 1 February 2021
(No difference)

Wang et al 2019b - Revision history

Scipediacontent: Scipediacontent moved page Draft Content 245975100 to Wang et al 2019b

Scipediacontent: Created page with " == Abstract == End-to-end speech synthesis is a promising approach that directly converts raw text to speech. Although it was shown that Tacotron2 outperforms classical pipe..."