-
Notifications
You must be signed in to change notification settings - Fork 4
/
index.html
35 lines (29 loc) · 1.95 KB
/
index.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
<html>
<head>
<link rel="stylesheet" href="./css/audio_samples.css">
<title>Audio samples for various CHiVE papers</title>
</head>
<body>
<center>
<div class="main_content">
<h1>Audio samples for CHiVE</h1>
Audio samples for various CHiVE (Clockwork Hierarchical Variational Autoencoder) papers.
<h2>Improving Prosody of RNN-based English Text-To-Speech Synthesis by Incorporating a BERT model</h2>
<div class="comment">
<a href="./chive-bert/index.html">These samples</a> go with <a href="http://www.interspeech2020.org/uploadfile/pdf/Thu-2-9-6.pdf" target="_blank">Improving Prosody of RNN-based English Text-To-Speech Synthesis by Incorporating a BERT model</a>, T. Kenter, M. Sharma, R. A. J. Clark, INTERSPEECH 2020.
<p>
We also released <a href="./chive-bert/dataset.html">the 'hard lines' evaluation set</a> as used in the paper.
</div>
<h2>Modelling Intonation in Spectrograms for Neural Vocoder based Text-to-Speech</h2>
<div class="comment">
<a href="./sp2020/index.html">These samples</a> go with <i>Modelling Intonation in Spectrograms for Neural Vocoder based Text-to-Speech</i>, V. Wan, J. Shen, H. Silen and R. A. J. Clark, Speech Prosody 2020.
</div>
<h2>CHiVE: Varying Prosody in Speech Synthesis with a Linguistically Driven Dynamic Hierarchical Conditional Variational Network</h2>
<div class="comment">
See <a href="https://google.github.io/chive-varying-prosody-icml-2019/" target="_blank">here</a> for audio samples that go with the original CHiVE paper
<i><a href="https://research.google/pubs/pub48239/" target="_blank">CHiVE: Varying Prosody in Speech Synthesis with a Linguistically Driven Dynamic Hierarchical Conditional Variational Network</a></i>, V. Wan, C-a Chan, T. Kenter, J. Vit and R. A. J. Clark, ICML 2019.
</div>
</div>
</center>
</body>
</html>