File size: 8,572 Bytes
5054992
 
0392634
5054992
 
 
0392634
 
 
5054992
0392634
5054992
0392634
5054992
 
 
 
 
0392634
5054992
 
 
 
 
 
 
 
 
0392634
5054992
 
0392634
 
 
 
 
 
baa9131
0392634
 
 
 
 
 
5054992
0392634
 
 
5054992
0392634
 
 
aaf5385
0392634
5054992
0392634
aaf5385
5054992
0392634
 
baa9131
0392634
 
 
 
baa9131
 
 
 
 
 
 
 
 
 
5054992
0392634
 
5054992
0392634
5054992
 
 
 
0392634
5054992
0392634
5054992
 
0392634
5054992
 
0392634
 
 
 
 
 
 
 
baa9131
 
 
 
 
 
 
 
81483a2
baa9131
 
 
 
5054992
 
 
 
0392634
 
5054992
0392634
5054992
 
0392634
 
baa9131
 
4686ffb
baa9131
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
d1b6c8e
baa9131
5054992
d1b6c8e
baa9131
5054992
0392634
5054992
 
0392634
 
 
 
 
 
 
 
5054992
0392634
 
5054992
 
0392634
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
5054992
 
 
0392634
5054992
 
0392634
baa9131
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
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
<!DOCTYPE html>
<html>

<head>
  <meta charset="utf-8">
  <meta name="description"
    content="Sidon is a fast, open-source multilingual speech restoration model for large-scale dataset restoration in TTS and spoken language modeling.">
  <meta name="keywords"
    content="speech restoration, dataset restoration, multilingual, TTS, spoken language models, vocoder, LoRA, w2v-BERT 2.0, HiFi-GAN">
  <meta name="viewport" content="width=device-width, initial-scale=1">
  <title>Sidon: Fast and Robust Open-Source Multilingual Speech Restoration</title>

  <link href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro" rel="stylesheet">

  <link rel="stylesheet" href="./static/css/bulma.min.css">
  <link rel="stylesheet" href="./static/css/bulma-carousel.min.css">
  <link rel="stylesheet" href="./static/css/bulma-slider.min.css">
  <link rel="stylesheet" href="./static/css/fontawesome.all.min.css">
  <link rel="stylesheet" href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
  <link rel="stylesheet" href="./static/css/index.css">
  <link rel="icon" href="./static/images/favicon.svg">

  <script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
  <script defer src="./static/js/fontawesome.all.min.js"></script>
  <script src="./static/js/bulma-carousel.min.js"></script>
  <script src="./static/js/bulma-slider.min.js"></script>
  <script src="./static/js/index.js"></script>
</head>

<body>

  <section class="hero">
    <div class="hero-body">
      <div class="container is-max-desktop">
        <div class="columns is-centered">
          <div class="column has-text-centered">
            <h1 class="title is-1 publication-title">Sidon: Fast and Robust Open-Source Multilingual Speech Restoration
              for Dataset Cleansing</h1>
            <div class="is-size-5 publication-authors">
              <span class="author-block">Wataru Nakata,</span>
              <span class="author-block">Yuki Saito,</span>
              <span class="author-block">Yota Ueda,</span>
              <span class="author-block">Hiroshi Saruwatari</span>
            </div>

            <div class="is-size-5 publication-authors">
              <span class="author-block">The University of Tokyo, Japan.</span>
            </div>

            <div class="column has-text-centered">
              <div class="publication-links">
                <span class="link-block">
                  <a href="https://arxiv.org/abs/2509.17052" target="_blank" class="external-link button is-normal is-rounded is-dark">
                    <span class="icon">
                      <i class="fas fa-file-pdf"></i>
                    </span>
                    <span>Paper</span>
                  </a>
                </span>
                <span class="link-block">
                  <a href="https://github.com/sarulab-speech/Sidon" target="_blank"
                    class="external-link button is-normal is-rounded is-dark">
                    <span class="icon">
                      <i class="fab fa-github"></i>
                    </span>
                    <span>Code </span>
                  </a>
                </span>
                <span class="link-block">
                  <a href="https://huggingface.co/spaces/sarulab-speech/sidon_demo_beta" target="_blank"
                    class="external-link button is-normal is-rounded is-dark">
                    <span class="icon">
                      🤗
                    </span>
                    <span>Live Demo</span>
                  </a>
                </span>
              </div>

            </div>
          </div>
        </div>
      </div>
    </div>
  </section>

  <!-- Teaser section removed for this project page. -->


  <!-- Results carousel removed for this project page. -->


  <section class="section">
    <div class="container is-max-desktop">
      <!-- Abstract. -->
      <div class="columns is-centered has-text-centered">
        <div class="column is-four-fifths">
          <h2 class="title is-3">Abstract</h2>
          <div class="content has-text-justified">
            <p>
              Large-scale text-to-speech (TTS) systems are limited by the scarcity of clean,
              multilingual recordings. We introduce <b>Sidon</b>, a fast, open-source
              speech restoration model that converts noisy in-the-wild speech into
              studio-quality speech and scales to dozens of languages. Sidon consists of
              two models: w2v-BERT 2.0 finetuned feature predictor to cleanse features from noisy speech
              and vocoder trained to synthesize restored speech from the cleansed features.
              Sidon achieves restoration performance comparable to Miipher: Google's internal speech restoration model with
              the aim of dataset cleansing for speech synthesis. Sidon is also computationally
              efficient, running up to 500× faster than real time on a
              single GPU. We further show that training a TTS model using a Sidon-cleansed automatic speech
              recognition corpus improves the quality of synthetic
              speech in a zero-shot setting. Code and model are released to
              facilitate reproducible dataset cleansing for the research community.
            </p>
          </div>
        </div>
      </div>
      <!--/ Abstract. -->
      <!-- Paper video section intentionally omitted. -->
    </div>
  </section>


  <section class="section">
    <div class="container is-max-desktop">
      <h2 class="title is-3 has-text-centered">Full Multilingual Results (FLEURS)</h2>
      <div class="content has-text-centered is-size-6">
        <p>The full multilingual evaluation table is large. It is hidden by default.</p>
      </div>
      <details id="fleurs-details" class="box">
        <summary class="is-size-5">Show results table</summary>
        <p class="is-size-7 has-text-grey">Loads an embedded page with all language-wise metrics.</p>
        <div id="fleurs-iframe-wrapper" style="margin-top: 0.75rem;">
          <iframe title="FLEURS Results"
                  src="full_result.html"
                  loading="lazy"
                  style="width: 100%; height: 70vh; border: 1px solid #e5e5e5; border-radius: 6px;"></iframe>
        </div>
      </details>
    </div>
  </section>


  <section class="section">
    <div class="container is-max-desktop">

      <h2 class="title is-3 has-text-centered">Multilingual Samples from FLEURS</h2>
      <div id="samples-multilingual-root" class="samples-root"></div>

      <h2 class="title is-3 has-text-centered" style="margin-top:2rem;">English Demo Samples from LibriTTS</h2>
      <div id="samples-english-root" class="samples-root"></div>
    </div>
  </section>


  <section class="section" id="BibTeX">
    <div class="container is-max-desktop content">
      <h2 class="title">BibTeX</h2>
      <pre><code>@inproceedings{sidon2026,
  author    = {Nakata, Wataru and Saito, Yuki and Ueda, Yota and Saruwatari, Hiroshi},
  title     = {Sidon: Fast and Robust Open-Source Multilingual Speech Restoration for Dataset Restoration},
  booktitle = {TBA},
  year      = {TBA}
}</code></pre>
    </div>
  </section>


  <footer class="footer">
    <div class="container">
      <div class="content has-text-centered">
        <a class="icon-link" target="_blank" href="#">
          <i class="fas fa-file-pdf"></i>
        </a>
        <a class="icon-link" href="https://huggingface.co/spaces/Wataru/SidonSamples" target="_blank"
          class="external-link">
          <i class="fab fa-github"></i>
        </a>
      </div>
      <div class="columns is-centered">
        <div class="column is-8">
          <div class="content">
            <p>
              This website is licensed under a <a rel="license" target="_blank"
                href="http://creativecommons.org/licenses/by-sa/4.0/">Creative
                Commons Attribution-ShareAlike 4.0 International License</a>.
            </p>
            <p>
              This means you are free to borrow the <a target="_blank"
                href="https://github.com/nerfies/nerfies.github.io">source code</a> of this website,
              we just ask that you link back to this page in the footer.
              Please remember to remove the analytics code included in the header of the website which
              you do not want on your website.
            </p>
          </div>
        </div>
      </div>
    </div>
  </footer>

</body>

</html>