R-Kentaren commited on
Commit
dcdfae3
·
verified ·
1 Parent(s): 1699200

Update index.html

Browse files
Files changed (1) hide show
  1. index.html +155 -761
index.html CHANGED
@@ -1,774 +1,168 @@
1
  <!DOCTYPE html>
2
  <html lang="en">
3
  <head>
4
- <meta charset="utf-8">
5
- <meta name="viewport" content="width=device-width, initial-scale=1, shrink-to-fit=no">
6
- <title>Zero-Shot Unsupervised and Text-Based Audio Editing Using DDPM Inversion</title>
7
- <link rel="shortcut icon" type="image/x-icon" href="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/favicon.ico">
8
- <meta property="og:image" content="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/teaser.png"/>
9
- <meta property="og:title" content="Zero-Shot Unsupervised and Text-Based Audio Editing Using DDPM Inversion" />
10
- <meta property="og:description" content="Zero-Shot Unsupervised and Text-Based Audio Editing Using DDPM Inversion" />
11
-
12
- <!-- Fonts -->
13
- <link rel="preconnect" href="https://fonts.googleapis.com">
14
- <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
15
- <link href="https://fonts.googleapis.com/css2?family=Poppins:wght@400;500;600;700;800&display=swap" rel="stylesheet">
16
-
17
- <!-- Bootstrap CSS -->
18
- <link href="https://cdn.jsdelivr.net/npm/bootstrap@5.3.3/dist/css/bootstrap.min.css" rel="stylesheet" integrity="sha384-QWTKZyjpPEjISv5WaRU9OFeRpok6YctnYmDr5pNlyT2bRjXh0JMhjY6hW+ALEwIH" crossorigin="anonymous">
19
-
20
- <!-- Tailwind CSS -->
21
- <script src="https://cdn.tailwindcss.com"></script>
22
-
23
- <!-- Academicons and Font Awesome -->
24
- <link rel="stylesheet" href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
25
- <script src="https://kit.fontawesome.com/ad96f96272.js" crossorigin="anonymous"></script>
26
-
27
- <!-- jQuery -->
28
- <script src="https://ajax.googleapis.com/ajax/libs/jquery/3.6.1/jquery.min.js"></script>
29
-
30
- <!-- Custom JavaScript -->
31
- <script type="text/javascript" src="functionality.js"></script>
32
 
33
- <style>
34
- body {
35
- @apply bg-gray-900 text-gray-100 font-poppins;
36
- }
37
- .navbar {
38
- @apply bg-gradient-to-r from-blue-600 to-blue-800 shadow-lg;
39
- }
40
- .btn-outline-primary {
41
- @apply border-blue-500 text-blue-500 hover:bg-blue-500 hover:text-white transition-colors duration-300;
42
- }
43
- .btn-back-to-top {
44
- @apply fixed bottom-8 right-8 bg-red-600 text-white rounded-full p-4 shadow-lg hover:bg-red-700 transition-colors duration-300;
45
- }
46
- .table {
47
- @apply bg-gray-800 text-gray-100 border-gray-700;
48
- }
49
- .table th, .table td {
50
- @apply border-gray-700;
51
- }
52
- .table thead th {
53
- @apply bg-gray-700 text-gray-200;
54
- }
55
- .audio-player {
56
- @apply w-full max-w-[200px] sm:max-w-[250px] md:max-w-[300px];
57
- }
58
- .shortaudio, .shorteraudio {
59
- @apply audio-player;
60
- }
61
- .section {
62
- @apply mb-8;
63
- }
64
- .anchor {
65
- @apply mt-12 mb-6;
66
- }
67
- .poppins-regular {
68
- @apply font-normal;
69
- }
70
- .poppins-medium {
71
- @apply font-medium;
72
- }
73
- @media (max-width: 768px) {
74
- h1 {
75
- @apply text-2xl;
76
- }
77
- h2 {
78
- @apply text-xl;
79
- }
80
- h3 {
81
- @apply text-lg;
82
- }
83
- .table-responsive {
84
- @apply overflow-x-auto;
85
- }
86
- .btn-lg {
87
- @apply text-sm px-3 py-1;
88
- }
89
- .audio-player {
90
- @apply max-w-[150px];
91
- }
92
- .navbar-brand {
93
- @apply text-lg;
94
- }
95
- }
96
- </style>
97
  </head>
98
  <body>
99
- <!-- Bootstrap JS -->
100
- <script src="https://cdn.jsdelivr.net/npm/bootstrap@5.3.3/dist/js/bootstrap.bundle.min.js" integrity="sha384-YvpcrYf0tY3lHB60NNkmXc5s9fDVZLESaAA55NDzOxhy9GkcIdslK1eN7N6jIeHz" crossorigin="anonymous"></script>
101
-
102
- <!-- Navigation Bar -->
103
- <nav id="navbar_top" class="navbar sticky-top navbar-expand-lg navbar-dark">
104
- <div class="container">
105
- <a class="navbar-brand poppins-medium" href="#">Zero-Shot Unsupervised and Text-Based Audio Editing</a>
106
- <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#main_nav" aria-controls="main_nav" aria-expanded="false" aria-label="Toggle navigation">
107
- <span class="navbar-toggler-icon"></span>
108
- </button>
109
- <div class="collapse navbar-collapse" id="main_nav">
110
- <ul class="navbar-nav ms-auto">
111
- <li class="nav-item"><a class="nav-link" href="#supsamples">Text-based Editing</a></li>
112
- <li class="nav-item"><a class="nav-link" href="#unsupsamples">Unsup. Editing</a></li>
113
- <li class="nav-item"><a class="nav-link" href="#supcomparisons">Text-Based Editing Comp.</a></li>
114
- <li class="nav-item"><a class="nav-link" href="#unsupcomparisons">Unsup. Editing Comp.</a></li>
115
- </ul>
116
- </div>
117
- </div>
118
- </nav>
119
-
120
- <div class="container-xl px-4 sm:px-6 lg:px-8">
121
- <!-- Back to Top Button -->
122
- <button type="button" class="btn btn-danger btn-floating btn-lg hidden" id="btn-back-to-top">
123
- <i class="fas fa-arrow-up"></i>
124
- </button>
125
-
126
- <!-- Header Section -->
127
- <div class="text-center py-8">
128
- <h1 class="poppins-medium text-3xl md:text-4xl lg:text-5xl mb-4">Zero-Shot Unsupervised and Text-Based Audio Editing Using DDPM Inversion</h1>
129
- <h4 class="poppins-regular text-xl md:text-2xl">ICML 2024</h4>
130
- <div class="flex flex-wrap justify-center gap-4 mt-4">
131
- <a href="https://www.linkedin.com/in/hilamanor/" target="_blank" rel="noopener noreferrer" class="text-2xl hover:text-blue-400 transition-colors">Hila Manor</a>
132
- <a href="https://tomer.net.technion.ac.il/" target="_blank" rel="noopener noreferrer" class="text-2xl hover:text-blue-400 transition-colors">Tomer Michaeli</a>
133
- </div>
134
- <p class="text-center text-lg md:text-xl mt-2">Technion - Israel Institute of Technology</p>
135
- <div class="flex flex-wrap justify-center gap-4 mt-6">
136
- <a class="btn btn-lg btn-outline-primary" href="https://arxiv.org/abs/2402.10009" target="_blank" rel="noopener noreferrer">
137
- <i class="ai ai-arxiv" aria-hidden="true"></i> ArXiv
138
- </a>
139
- <a class="btn btn-lg btn-outline-primary" href="https://github.com/HilaManor/AudioEditingCode/" target="_blank" rel="noopener noreferrer">
140
- <i class="fa fa-github" aria-hidden="true"></i> Code
141
- </a>
142
- <a class="btn btn-lg btn-outline-primary" href="https://youtu.be/lBnldOQVbS4" target="_blank" rel="noopener noreferrer">
143
- <i class="fa fa-youtube" aria-hidden="true"></i> Presentation
144
- </a>
145
- <a class="btn btn-lg btn-outline-primary" href="https://huggingface.co/spaces/hilamanor/audioEditing/" target="_blank" rel="noopener noreferrer">
146
- 🤗 Text-Based Space
147
- </a>
148
- </div>
149
- </div>
150
-
151
- <!-- Abstract and Teaser Image -->
152
- <div class="text-center py-8">
153
- <img alt="Teaser" id="teaser" src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/teaser.png" class="max-w-full w-full md:w-3/4 lg:w-2/3 mx-auto rounded-lg shadow-lg"/>
154
- <h1 class="poppins-regular text-2xl md:text-3xl mt-6">Abstract</h1>
155
- <p class="text-base md:text-lg leading-relaxed mt-4">
156
- Editing signals using large pre-trained models, in a zero-shot manner, has recently seen rapid advancements in the image domain. However, this wave has yet to reach the audio domain.
157
- In this paper, we explore two zero-shot editing techniques for audio signals, which use DDPM inversion on pre-trained diffusion models. The first, adopted from the image domain, allows text-based editing.
158
- The second, is a novel approach for discovering semantically meaningful editing directions without supervision.
159
- When applied to music signals, this method exposes a range of musically interesting modifications, from controlling the participation of specific instruments to improvisations on the melody.
160
- </p>
161
- </div>
162
-
163
- <hr class="border-gray-700 my-8">
164
-
165
- <!-- Video Overview -->
166
- <div class="text-center py-8">
167
- <h1 class="poppins-regular text-2xl md:text-3xl">Video Overview</h1>
168
- <p class="text-base md:text-lg mt-2">For people in a hurry. Images were generated by <a href="https://openai.com/dall-e-2" class="text-blue-400 hover:underline">DALL-E 2</a> and <a href="https://www.bing.com/copilot" class="text-blue-400 hover:underline">Copilot</a>.</p>
169
- <video controls class="w-full max-w-[400px] mx-auto rounded-lg shadow-lg mt-4">
170
- <source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/overview.mp4" type="video/mp4">
171
- Your browser does not support the video tag.
172
- </video>
173
- </div>
174
 
175
- <hr class="border-gray-700 my-8">
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
176
 
177
- <!-- Samples of Editing -->
178
- <div class="text-center py-8">
179
- <h1 class="poppins-regular text-2xl md:text-3xl">1. Samples of Editing</h1>
180
- <p class="text-base md:text-lg mt-2">We present samples of audio editing using our proposed methods. The samples are organized into two sections: text-based editing and unsupervised editing.</p>
 
 
 
 
 
 
 
181
 
182
- <!-- Text-Based Editing -->
183
- <h2 id="supsamples" class="anchor poppins-regular text-xl md:text-2xl flex items-center justify-center">
184
- <button type="button" id="variousSupSamplesBtn" onclick="toggleCollapseArrow('variousSupSamplesBtn')" class="btn btn-light mr-2" data-bs-toggle="collapse" data-bs-target="#variousSupSamples" aria-expanded="true" aria-controls="variousSupSamples"><i class="fas fa-chevron-down"></i></button>
185
- 1.1. Samples of Text-Based Editing
186
- </h2>
187
- <div class="container show" id="variousSupSamples">
188
- <div class="table-responsive">
189
- <table class="table table-sm">
190
- <thead>
191
- <tr class="text-center">
192
- <th scope="col">#</th>
193
- <th scope="col">Source Prompt</th>
194
- <th scope="col">Target Prompt</th>
195
- <th scope="col">Original Audio</th>
196
- <th scope="col">Edited Audio</th>
197
- <th scope="col">Edit T<sub>start</sub></th>
198
- </tr>
199
- </thead>
200
- <tbody>
201
- <tr>
202
- <th scope="row">1</th>
203
- <td>A recording of a <b>sneaky jazz</b> song.</td>
204
- <td>A recording of a <b>tense classical</b> music score.</td>
205
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBFreeJazz.mp3" type="audio/mp4"></audio></td>
206
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/sup_samples/orchestra_MDDBFreeJazz_ours_90.mp3" type="audio/mp4"></audio></td>
207
- <td class="text-center align-middle">110</td>
208
- </tr>
209
- <tr>
210
- <th scope="row">2</th>
211
- <td>A recording of a <b>hard rock</b> song.</td>
212
- <td>A recording of a <b>jazz</b> song.</td>
213
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBZeppelin.mp3" type="audio/mp4"></audio></td>
214
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/sup_samples/jazz_MDDBZeppelin_ours_100.mp3" type="audio/mp4"></audio></td>
215
- <td class="text-center align-middle">100</td>
216
- </tr>
217
- <tr>
218
- <th scope="row">3</th>
219
- <td>A recording of a happy upbeat <b>classical music piece</b>.</td>
220
- <td>A recording of a happy upbeat <b>arcade game soundtrack</b>.</td>
221
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBVivaldi.mp3" type="audio/mp4"></audio></td>
222
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/sup_samples/arcade_MDDBVivaldi_ours_100.mp3" type="audio/mp4"></audio></td>
223
- <td class="text-center align-middle">100</td>
224
- </tr>
225
- <tr>
226
- <th scope="row">4</th>
227
- <td><b>Trumpets</b> playing alongside a piano, bass and drums in an upbeat old-timey cool <b>jazz</b> song.</td>
228
- <td><b>A banjo</b> playing alongside a piano, bass and drums in an upbeat old-timey cool <b>country</b> song.</td>
229
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBModalJazz.mp3" type="audio/mp4"></audio></td>
230
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/sup_samples/country_MDDBModalJazz_ours_110.mp3" type="audio/mp4"></audio></td>
231
- <td class="text-center align-middle">90</td>
232
- </tr>
233
- <tr>
234
- <th scope="row">5</th>
235
- <td class="text-center align-middle">——</td>
236
- <td>A recording of a dark techno song.</td>
237
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBBeethoven.mp3" type="audio/mp4"></audio></td>
238
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/sup_samples/techno_MDDBBeethoven_ours_110.mp3" type="audio/mp4"></audio></td>
239
- <td class="text-center align-middle">90</td>
240
- </tr>
241
- <tr class="text-center align-middle">
242
- <td colspan="6">
243
- <button class="btn btn-outline-secondary" type="button" id="supExamplesHearMore1Btn" onclick="hearMore('supExamplesHearMore1Btn', 6, 13)">
244
- <i class="fa fa-plus-square-o" aria-hidden="true"></i> I want to hear more!
245
- </button>
246
- </td>
247
- </tr>
248
- <tr hidden>
249
- <th scope="row">6</th>
250
- <td>A recording of a dramatic <b>epic Chinese</b> piece.</td>
251
- <td>A recording of a dramatic <b>heavy metal</b> piece.</td>
252
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBChineseDrama.mp3" type="audio/mp4"></audio></td>
253
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/sup_samples/metal_MDDBChineseDrama_ours_160.mp3" type="audio/mp4"></audio></td>
254
- <td class="text-center align-middle">40</td>
255
- </tr>
256
- <tr hidden>
257
- <th scope="row">7</th>
258
- <td>A recording of <b>a rock song</b>.</td>
259
- <td>A recording of <b>Arabic music</b>.</td>
260
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBBritpop.mp3" type="audio/mp4"></audio></td>
261
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/sup_samples/arabic_MDDBBritpop_ours_90.mp3" type="audio/mp4"></audio></td>
262
- <td class="text-center align-middle">110</td>
263
- </tr>
264
- <tr hidden>
265
- <th scope="row">8</th>
266
- <td class="text-center align-middle">——</td>
267
- <td>A recording of a funky hip hop song.</td>
268
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBFunkJazz.mp3" type="audio/mp4"></audio></td>
269
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/sup_samples/hiphop_MDDBFunkJazz_ours_90.mp3" type="audio/mp4"></audio></td>
270
- <td class="text-center align-middle">110</td>
271
- </tr>
272
- <tr hidden>
273
- <th scope="row">9</th>
274
- <td>A high quality recording of <b>wind instruments and strings</b> playing.</td>
275
- <td>A high quality recording of <b>a piano</b> playing.</td>
276
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBBeethoven.mp3" type="audio/mp4"></audio></td>
277
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/sup_samples/instrument_MDDBBeethoven_ours_130.mp3" type="audio/mp4"></audio></td>
278
- <td class="text-center align-middle">70</td>
279
- </tr>
280
- <tr hidden>
281
- <th scope="row">10</th>
282
- <td>A recording of an upbeat <b>gospel</b> song.</td>
283
- <td>A recording of an upbeat <b>techno</b> song.</td>
284
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBGospel.mp3" type="audio/mp4"></audio></td>
285
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/sup_samples/techno_MDDBGospel_ours_100.mp3" type="audio/mp4"></audio></td>
286
- <td class="text-center align-middle">100</td>
287
- </tr>
288
- <tr hidden>
289
- <th scope="row">11</th>
290
- <td>A recording of a happy upbeat song in a <b>Latin jazz</b> style.</td>
291
- <td>A recording of a happy upbeat song in a <b>retro arcade game soundtrack</b> style.</td>
292
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBLatinJazz.mp3" type="audio/mp4"></audio></td>
293
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/sup_samples/arcade_MDDBLatinJazz_ours_110.mp3" type="audio/mp4"></audio></td>
294
- <td class="text-center align-middle">90</td>
295
- </tr>
296
- <tr hidden>
297
- <th scope="row">12</th>
298
- <td class="text-center align-middle">——</td>
299
- <td>A recording of an upbeat cool jazz song.</td>
300
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBPunk.mp3" type="audio/mp4"></audio></td>
301
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/sup_samples/jazz_MDDBPunk_ours_110.mp3" type="audio/mp4"></audio></td>
302
- <td class="text-center align-middle">90</td>
303
- </tr>
304
- <tr class="text-center align-middle" hidden>
305
- <td colspan="6">
306
- <button class="btn btn-outline-secondary" type="button" id="supExamplesHearMore2Btn" onclick="hearMore('supExamplesHearMore2Btn', 14, 19)">
307
- <i class="fa fa-plus-square-o" aria-hidden="true"></i> I want to hear even more!
308
- </button>
309
- </td>
310
- </tr>
311
- <tr hidden>
312
- <th scope="row">13</th>
313
- <td>A recording of an <b>old rock</b> song.</td>
314
- <td>A recording of an <b>techno</b> song.</td>
315
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBRockabilly.mp3" type="audio/mp4"></audio></td>
316
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/sup_samples/techno_MDDBRockabilly_ours_110.mp3" type="audio/mp4"></audio></td>
317
- <td class="text-center align-middle">90</td>
318
- </tr>
319
- <tr hidden>
320
- <th scope="row">14</th>
321
- <td>Chinese strings, flutes, and harps playing an <b>upbeat</b> piece.</td>
322
- <td>Chinese strings, flutes, and harps playing an <b>somber</b> piece.</td>
323
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBChineseChaoZhou.mp3" type="audio/mp4"></audio></td>
324
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/sup_samples/tone_MDDBChineseChaoZhou_ours_120.mp3" type="audio/mp4"></audio></td>
325
- <td class="text-center align-middle">80</td>
326
- </tr>
327
- <tr hidden>
328
- <th scope="row">15</th>
329
- <td class="text-center align-middle">——</td>
330
- <td>A recording of an upbeat arcade game soundtrack.</td>
331
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBBebopJazz.mp3" type="audio/mp4"></audio></td>
332
- <td><audio preload="auto" controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/sup_samples/arcade_MDDBBebopJazz_ours_120.mp3" type="audio/mp4"></audio></td>
333
- <td class="text-center align-middle">80</td>
334
- </tr>
335
- <tr hidden>
336
- <th scope="row">16</th>
337
- <td>A high quality recording of a <b>cat meowing</b>.</td>
338
- <td>A high quality recording of a <b>dog barking</b>.</td>
339
- <td><audio controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/CuteCat_10sec.mp3" type="audio/mp4"></audio></td>
340
- <td><audio controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/sup_samples/CuteCat_10sec_ours_50.mp3" type="audio/mp4"></audio></td>
341
- <td class="text-center">150</td>
342
- </tr>
343
- <tr hidden>
344
- <th scope="row">17</th>
345
- <td>A high quality recording of a <b>dog barking</b> a lot.</td>
346
- <td>A high quality recording of a <b>gun shooting</b> a lot.</td>
347
- <td><audio controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/DogBarking_10sec.mp3" type="audio/mp4"></audio></td>
348
- <td><audio controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/sup_samples/DogBarking_10sec_ours_100.mp3" type="audio/mp4"></audio></td>
349
- <td class="text-center">100</td>
350
- </tr>
351
- <tr hidden>
352
- <th scope="row">18</th>
353
- <td>A <b>kid talking loudly</b>.</td>
354
- <td>A <b>rooster crowing</b>.</td>
355
- <td><audio controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/Shouting.mp3" type="audio/mp4"></audio></td>
356
- <td><audio controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/sup_samples/Shouting_ours_90.mp3" type="audio/mp4"></audio></td>
357
- <td class="text-center">110</td>
358
- </tr>
359
- </tbody>
360
- </table>
361
- </div>
362
- <div class="mt-6">
363
- <div class="flex justify-center">
364
- <button class="btn btn-lg btn-warning" type="button" data-bs-toggle="collapse" data-bs-target="#huggingface_demo" aria-expanded="false" aria-controls="huggingface_demo">🤗 Space Demo</button>
365
- </div>
366
- <div id="huggingface_demo" class="collapse mt-4">
367
- <div class="card-body">
368
- <iframe src="https://hilamanor-audioediting.hf.space" frameborder="0" class="w-full h-[600px] rounded-lg shadow-lg"></iframe>
369
- </div>
370
- </div>
371
- </div>
372
- </div>
373
 
374
- <!-- Unsupervised Editing -->
375
- <h2 id="unsupsamples" class="anchor poppins-regular text-xl md:text-2xl">1.2. Samples of Unsupervised Editing</h2>
376
- <p class="text-base md:text-lg mt-2">For the unsupervised editing, we split the samples into two sections.<br>The first section (Strength changes) shows how the same direction applied with different strengths changes the audio sequentially.<br>The second section (PC direction changes) shows how removing or adding a direction removes or adds a concept.</p>
377
-
378
- <h3 class="poppins-regular text-lg md:text-xl flex items-center justify-center">
379
- <button type="button" id="variousUnsupSamplesStrengthBtn" onclick="toggleCollapseArrow('variousUnsupSamplesStrengthBtn')" class="btn btn-light mr-2" data-bs-toggle="collapse" data-bs-target="#variousUnsupSamplesStrength" aria-expanded="true" aria-controls="variousUnsupSamplesStrength"><i class="fas fa-chevron-down"></i></button>
380
- 1.2.1. Various Samples (Strength changes)
381
- </h3>
382
- <div class="container show" id="variousUnsupSamplesStrength">
383
- <div class="table-responsive">
384
- <table class="table table-sm">
385
- <thead>
386
- <tr class="text-center">
387
- <th scope="col">#</th>
388
- <th scope="col">Inversion Prompt</th>
389
- <th scope="col">Original Audio</th>
390
- <th scope="col">Edited Audio +PC</th>
391
- <th scope="col">Edited Audio +2PC</th>
392
- <th scope="col">PC Interpretation</th>
393
- <th scope="col">Edit Parameters</th>
394
- </tr>
395
- </thead>
396
- <tbody>
397
- <tr>
398
- <th scope="row">1</th>
399
- <td>A high quality recording of flutes and a trumpet playing.</td>
400
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBLatinJazz_8secs.mp3" type="audio/mp4"></audio></td>
401
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBLatinJazz_ours_200--1_spts80_pcs_a-1.mp3" type="audio/mp4"></audio></td>
402
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBLatinJazz_ours_200--1_spts80_pcs_a-2.mp3" type="audio/mp4"></audio></td>
403
- <td>Melody change</td>
404
- <td class="text-center">t'∈[200, -1]<br>Specific t=80 used<br>PCs 1+2+3</td>
405
- </tr>
406
- <tr>
407
- <th scope="row">2</th>
408
- <td>A recording of a calm country song.</td>
409
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBCountry1_8secs.mp3" type="audio/mp4"></audio></td>
410
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBCountry1_ours_150--1_spts115_pcs_a1.mp3" type="audio/mp4"></audio></td>
411
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBCountry1_ours_150--1_spts115_pcs_a2.mp3" type="audio/mp4"></audio></td>
412
- <td>Remove singer</td>
413
- <td class="text-center">t'∈[150, -1]<br>Specific t=115 used<br>PCs 1+2+3</td>
414
- </tr>
415
- <tr>
416
- <th scope="row">3</th>
417
- <td class="text-center">—</td>
418
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBReggae_8secs.mp3" type="audio/mp4"></audio></td>
419
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBReggae_ours_150--1_spts80_pcs_a-1.mp3" type="audio/mp4"></audio></td>
420
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBReggae_ours_150--1_spts80_pcs_a-2.mp3" type="audio/mp4"></audio></td>
421
- <td>Just drums</td>
422
- <td class="text-center">t'∈[150, -1]<br>Specific t=80 used<br>PCs 1+2+3</td>
423
- </tr>
424
- <tr>
425
- <th scope="row">4</th>
426
- <td>A recording of a scary classical music piece.</td>
427
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBInTheHalloftheMountainKing.mp3" type="audio/mp4"></audio></td>
428
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBInTheHalloftheMountainKing_ours_150-50_spts95_pcs_a1.mp3" type="audio/mp4"></audio></td>
429
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBInTheHalloftheMountainKing_ours_150-50_spts95_pcs_a2.mp3" type="audio/mp4"></audio></td>
430
- <td>Melody change</td>
431
- <td class="text-center">t'∈[150, 50]<br>Specific t=95 used<br>PCs 1+2+3</td>
432
- </tr>
433
- <tr>
434
- <th scope="row">5</th>
435
- <td>A trumpet and a saxophone playing a cool jazz melody, with an accompaniment of a piano, bass and drums.</td>
436
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBCoolJazz.mp3" type="audio/mp4"></audio></td>
437
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBCoolJazz_ours_135-95_pcs_a20.mp3" type="audio/mp4"></audio></td>
438
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBCoolJazz_ours_135-95_pcs_a40.mp3" type="audio/mp4"></audio></td>
439
- <td>Melody change</td>
440
- <td class="text-center">t'∈[135, 95]<br>PCs 1+2+3</td>
441
- </tr>
442
- <tr class="text-center align-middle">
443
- <td colspan="7">
444
- <button class="btn btn-outline-secondary" type="button" id="unsupExamplesStrengthHearMoreBtn" onclick="hearMore('unsupExamplesStrengthHearMoreBtn', 6, 10)">
445
- <i class="fa fa-plus-square-o" aria-hidden="true"></i> I want to hear more!
446
- </button>
447
- </td>
448
- </tr>
449
- <tr hidden>
450
- <th scope="row">6</th>
451
- <td>A high quality recording of wind instruments and strings playing.</td>
452
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBBeethoven.mp3" type="audio/mp4"></audio></td>
453
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBBeethoven_ours_135-95_pcs_a-20.mp3" type="audio/mp4"></audio></td>
454
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBBeethoven_ours_135-95_pcs_a-40.mp3" type="audio/mp4"></audio></td>
455
- <td>Melody change</td>
456
- <td class="text-center">t'∈[135, 95]<br>PCs 1+2+3</td>
457
- </tr>
458
- <tr hidden>
459
- <th scope="row">7</th>
460
- <td>A strings section playing classical music.</td>
461
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBVivaldi.mp3" type="audio/mp4"></audio></td>
462
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBVivaldi_ours_95-80_pcs_a-20.mp3" type="audio/mp4"></audio></td>
463
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBVivaldi_ours_95-80_pcs_a-40.mp3" type="audio/mp4"></audio></td>
464
- <td>Minor melody changes</td>
465
- <td class="text-center">t'∈[95, 80]<br>PCs 1</td>
466
- </tr>
467
- <tr hidden>
468
- <th scope="row">8</th>
469
- <td>A high quality recording of a woman singing while a guitar and drums play in the background.</td>
470
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBRockabilly_8secs.mp3" type="audio/mp4"></audio></td>
471
- </td><td><audio class="shortaudio" controls src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBRockabilly_ours_200--1_spts65_pcs_a-1.mp3"></td>
472
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBRockabilly_ours_200--1_spts65_pcs_a-2.mp3"></audio></td>
473
- <td>Instrument change</td>
474
- <td>t'∈[200, -1]<br>Specific t=65 used<br>PCs 1</td>
475
- </tr>
476
- <tr hidden>
477
- <th scope="row">9</th>
478
- <td>A high quality recording of wind instruments and strings playing.</td>
479
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBBeethoven.mp3"></audio></td>
480
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBBeethoven_ours_200--1_spts135_pc3_a1.mp3"></audio></td>
481
- <td><audio class="shortaudio" controls><source src="/driver://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDBBeethoven_ours_200--1_spts135_pc3_a2.mp3"></audio></td>
482
- <td>Melody changes</td>
483
- <td>t'∈[200, -1]<br>Specific t=135 used<br>PC #3</td>
484
- </tr>
485
- <tr hidden>
486
- <th scope="row">10</th>
487
- <td>A high quality recording of a rock band playing an upbeat rock song.</td>
488
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBritpop_8secs.mp3"></audio></td>
489
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBritpop_ours_200--1_spts95_pcs_a1.mp3"></audio></td>
490
- <td><audio class="shortaudio" controls><source src="https://github.com/hilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBritpop_ours_200--1_spts95_pcs_a2.mp3"></td>
491
- <td>Drums style change</td>
492
- <td>t'∈[200, -1]<br>Specific t=95 used<br>PCs 1+3</td>
493
- </tbody>
494
- </table>
495
- </div>
496
- </div>
497
- <h3 id="unsup-singlepc" class="anchor poppins-regular">1.2. Various Samples (PC direction changes)</h3>
498
- <div class="container">
499
- <div class="table-responsive">
500
- <table class="table">
501
- <thead>
502
- <tr class="text-center">
503
- <th scope="col">#</th>
504
- <th>Inversion Prompt</th>
505
- <th scope="col">Edited Audio -γPC</th>
506
- <th scope="col">Original Audio</th>
507
- <th scope="col">Edited Audio +γPC</th>
508
- <th scope="col">PC Interpretation</th>
509
- <th scope="col">Edit Parameters</th>
510
- </tr>
511
- </thead>
512
- <tbody>
513
- <tr>
514
- <th scope="row">1</th>
515
- <td>A high quality recording of a man singing and drums, guitar and bass playing a song, and later a woman singing.</td>
516
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDBDBeatles_ours_115-80_pc1_a-40.mp3"></audio></td>
517
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBBeatles.mp3"></audio></td>
518
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBBeatles_ours_115-80_pc1_a40.mp3"></audio></td>
519
- <td>Lead Guitar/Singer emphasis</td>
520
- <td>t'∈[115, 80]<br>PC #1</td>
521
- </tr>
522
- <tr>
523
- <th scope="row">2</th>
524
- <td>A high quality recording of a man singing and drums, guitar and bass playing a song, and later a woman singing.</td>
525
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBBeatles_ours_115-80_pc2_a-40.mp3"></audio></td>
526
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDDBBeatles.mp3"></audio></td>
527
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBBeatles_ours_115-80_pc2_a40.mp3"></audio></td>
528
- <td>Singers/Drums emphasis</td>
529
- <td>t'∈[115, 80]<br>PC #2</td>
530
- </tr>
531
- <tr>
532
- <th scope="row">3</th>
533
- <td>A high quality recording of a man singing with a rock band accompaniment.</td>
534
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBHendrix_ours_200--1_spts80_pc1_a-2.mp3"></audio></td>
535
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBHendrix_8secs.mp3"></audio></td>
536
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBHendrix_ours_200--1_spts80_pc1_a2.mp3"></audio></td>
537
- <td>Drum-beats style</td>
538
- <td>t'∈[200, -1]<br>Specific t=80 used<br>PC #1</td>
539
- </tr>
540
- <tr class="text-center align-middle">
541
- <td colspan="7">
542
- <button class="btn btn-outline-secondary" type="button" id="unsupExamplesDirectionHearMoreBtn" onclick="hearMore('unsupExamplesDirectionHearMoreBtn', 4, 6)">
543
- <i class="fa fa-plus-square-o" aria-hidden="true"></i> I want to hear more!
544
- </button>
545
- </td>
546
- </tr>
547
- <tr hidden>
548
- <th scope="row">4</th>
549
- <td>A recording of rhythmic clapping, a woman singing, and drums and guitar playing.</td>
550
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBGospel_ours_150--1_spts120_pc3_a-2.mp3"></audio></td>
551
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBGospel_8secs.mp3"></audio></td>
552
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBGospel_ours_150--1_spts120_pc3_a2.mp3"></audio></td>
553
- <td>Vibrato strength</td>
554
- <td>t'∈[150, -1]<br>Specific t=120 used<br>PC #3</td>
555
- </tr>
556
- <tr hidden>
557
- <th scope="row">5</th>
558
- <td>A recording of an old timey rock song from the sixties.</td>
559
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBRock_ours_200--1_spts65_pcs_a-2.mp3"></audio></td>
560
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBRock_8secs.mp3"></audio></td>
561
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/MDDBRock_ours_200--1_spts65_pcs_a2.mp3"></audio></td>
562
- <td>Guitar/Singer emphasis</td>
563
- <td>t'∈[200, -1]<br>Specific t=65 used<br>PCs t</td>
564
- </tr>
565
- <tr hidden>
566
- <th scope="row">6</th>
567
- <td class="text-center">—</td>
568
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/ManSpeakingBGNoise_ours_115-95_pc1_a-40.mp3"></audio></td>
569
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/ManSpeakingBGNoise.mp3"></audio></td>
570
- <td><audio class="shortaudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/unsup_samples/ManSpeakingBGNoise_ours_115-95_pc1_a40.mp3"></audio></td>
571
- <td>Isolate Woman/Man</td>
572
- <td>t'∈[115, 95]<br>PC #1</td>
573
- </tr>
574
- </tbody>
575
- </table>
576
- </div>
577
- </div>
578
- </div>
579
 
580
- <hr class="border-gray-700">
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
581
 
582
- <!-- Comparisons to Other Methods -->
583
- <div class="text-center py-8">
584
- <h1 class="poppins-regular text-2xl md:text-3xl">2. Comparisons to Other Methods</h1>
 
585
 
586
- <h2 id="supcomparisons" class="anchor poppins-regular text-xl md:text-2xl">2.1. Comparisons of Text-Based Editing</h2>
587
- <h3 class="poppins-regular text-lg md:text-xl flex items-center justify-center">
588
- <button type="button" id="supComparisonsMusicSamplesBtn" onclick="toggleCollapseArrow('supComparisonsMusicSamplesBtn')" class="btn btn-light mr-2" data-bs-toggle="collapse" data-bs-target="#supComparisonsMusicSamples" aria-expanded="true" aria-controls="supComparisonsMusicSamples"><i class="fas fa-chevron-down"></i></button>
589
- 2.1.2. Music Samples
590
- </h3>
591
- <div class="container show" id="supComparisonsMusicSamples">
592
- <div class="table-responsive">
593
- <table class="table table-sm">
594
- <thead>
595
- <tr class="text-center">
596
- <th scope="col">#</th>
597
- <th scope="col">Source Prompt</th>
598
- <th scope="col">Target Prompt</th>
599
- <th scope="col">Original Audio</th>
600
- <th scope="col">Ours</th>
601
- <th scope="col">SDEdit<br>T<sub>start</sub>=100<br>T<sub>start</sub>=70<br>T<sub>start</sub>=40</th>
602
- <th scope="col">MusicGen</th>
603
- <th scope="col">DDIM Editing</th>
604
- </tr>
605
- </thead>
606
- <tbody>
607
- <tr class="align-top">
608
- <th scope="row">1</th>
609
- <td class="align-top">A recording of <b>a rock song</b>.</td>
610
- <td class="align-top">A recording of <b>Arabic music</b>.</td>
611
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/orig/MDDBBritpop.mp3" type="audio/mp4"></audio></td>
612
- <td class="text-center align-middle"><br>
613
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sup_samples/arabic_MDDBBritpop_ours_90.mp3" type="audio/mp4"></audio>
614
- <br>T<sub>start</sub>=110</td>
615
- <td class="text-center align-middle"><br>
616
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/arabic_MDDBBritpop_sdedit_100.mp3" type="audio/mp4"></audio>
617
- <br>
618
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/arabic_MDDBBritpop_sdedit_130.mp3" type="audio/mp4"></audio>
619
- <br>
620
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/arabic_MDDBBritpop_sdedit_160.mp3" type="audio/mp4"></audio>
621
- </td>
622
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/musicgen/arabic_MDDBBritpop_musicgen.mp3" type="audio/mp4"></audio></td>
623
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/ddim/arabic_MDDBBritpop_ddim.mp3" type="audio/mp4"></audio></td>
624
- </tr>
625
- <tr class="align-top">
626
- <th scope="row">2</th>
627
- <td class="align-top">A recording of an <b>upbeat rock song</b>.</td>
628
- <td class="align-top">A recording of an <b>arcade game soundtrack</b>.</td>
629
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/orig/MDDBBeatles.mp3" type="audio/mp4"></audio></td>
630
- <td class="text-center align-middle"><br>
631
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sup_samples/arcade_MDDBBeatles_ours_100.mp3" type="audio/mp4"></audio>
632
- <br>T<sub>start</sub>=100</td>
633
- <td class="text-center align-middle"><br>
634
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/arcade_MDDBBeatles_sdedit_100.mp3" type="audio/mp4"></audio>
635
- <br>
636
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/arcade_MDDBBeatles_sdedit_130.mp3" type="audio/mp4"></audio>
637
- <br>
638
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/arcade_MDDBBeatles_sdedit_160.mp3" type="audio/mp4"></audio>
639
- </td>
640
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/musicgen/arcade_MDDBBeatles_musicgen.mp3" type="audio/mp4"></audio></td>
641
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/ddim/arcade_MDDBBeatles_ddim.mp3" type="audio/mp4"></audio></td>
642
- </tr>
643
- <tr class="align-top">
644
- <th scope="row">3</th>
645
- <td class="align-top text-center">—</td>
646
- <td class="align-top">A recording of a dark techno song.</td>
647
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/orig/MDDBBeethoven.mp3" type="audio/mp4"></audio></td>
648
- <td class="text-center align-middle"><br>
649
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sup_samples/techno_MDDBBeethoven_ours_110.mp3" type="audio/mp4"></audio>
650
- <br>T<sub>start</sub>=90</td>
651
- <td class="text-center align-middle"><br>
652
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/techno_MDDBBeethoven_sdedit_100.mp3" type="audio/mp4"></audio>
653
- <br>
654
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/techno_MDDBBeethoven_sdedit_130.mp3" type="audio/mp4"></audio>
655
- <br>
656
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/techno_MDDBBeethoven_sdedit_160.mp3" type="audio/mp4"></audio>
657
- </td>
658
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/musicgen/techno_MDDBBeethoven_musicgen.mp3" type="audio/mp4"></audio></td>
659
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/ddim/techno_MDDBBeethoven_ddim.mp3" type="audio/mp4"></audio></td>
660
- </tr>
661
- <tr class="align-top">
662
- <th scope="row">4</th>
663
- <td class="align-top">A high quality recording of <b>wind instruments and strings</b> playing.</td>
664
- <td class="align-top">A high quality recording of <b>a piano</b> playing.</td>
665
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/orig/MDDBBeethoven.mp3" type="audio/mp4"></audio></td>
666
- <td class="text-center align-middle"><br>
667
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sup_samples/instrument_MDDBBeethoven_ours_130.mp3" type="audio/mp4"></audio>
668
- <br>T<sub>start</sub>=70</td>
669
- <td class="text-center align-middle"><br>
670
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/instrument_MDDBBeethoven_sdedit_100.mp3" type="audio/mp4"></audio>
671
- <br>
672
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/instrument_MDDBBeethoven_sdedit_130.mp3" type="audio/mp4"></audio>
673
- <br>
674
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/instrument_MDDBBeethoven_sdedit_160.mp3" type="audio/mp4"></audio>
675
- </td>
676
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/musicgen/instrument_MDDBBeethoven_musicgen.mp3" type="audio/mp4"></audio></td>
677
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/ddim/instrument_MDDBBeethoven_ddim.mp3" type="audio/mp4"></audio></td>
678
- </tr>
679
- <tr class="text-center align-middle">
680
- <td colspan="8">
681
- <button class="btn btn-outline-secondary" type="button" id="supExamplesCompHearMoreBtn1" onclick="hearMore('supExamplesCompHearMoreBtn1', 5, 11)">
682
- <i class="fa fa-plus-square-o" aria-hidden="true"></i> I want to hear more!
683
- </button>
684
- </td>
685
- </tr>
686
- <tr class="align-top" hidden>
687
- <th scope="row">5</th>
688
- <td class="align-top text-center">—</td>
689
- <td class="align-top">A recording of an upbeat cool jazz song.</td>
690
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/orig/MDDBPunk.mp3" type="audio/mp4"></audio></td>
691
- <td class="text-center align-middle"><br>
692
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sup_samples/jazz_MDDBPunk_ours_110.mp3" type="audio/mp4"></audio>
693
- <br>T<sub>start</sub>=90</td>
694
- <td class="text-center align-middle"><br>
695
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/jazz_MDDBPunk_sdedit_100.mp3" type="audio/mp4"></audio>
696
- <br>
697
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/jazz_MDDBPunk_sdedit_130.mp3" type="audio/mp4"></audio>
698
- <br>
699
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/jazz_MDDBPunk_sdedit_160.mp3" type="audio/mp4"></audio>
700
- </td>
701
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/musicgen/jazz_MDDBPunk_musicgen.mp3" type="audio/mp4"></audio></td>
702
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/ddim/jazz_MDDBPunk_ddim.mp3" type="audio/mp4"></audio></td>
703
- </tr>
704
- <tr class="align-top" hidden>
705
- <th scope="row">6</th>
706
- <td class="align-top">A recording of an <b>old rock</b> song.</td>
707
- <td class="align-top">A recording of an <b>techno</b> song.</td>
708
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/orig/MDDBRockabilly.mp3" type="audio/mp4"></audio></td>
709
- <td class="text-center align-middle"><br>
710
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sup_samples/techno_MDDBRockabilly_ours_110.mp3" type="audio/mp4"></audio>
711
- <br>T<sub>start</sub>=90</td>
712
- <td class="text-center align-middle"><br>
713
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/techno_MDDBRockabilly_sdedit_100.mp3" type="audio/mp4"></audio>
714
- <br>
715
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/techno_MDDBRockabilly_sdedit_130.mp3" type="audio/mp4"></audio>
716
- <br>
717
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/techno_MDDBRockabilly_sdedit_160.mp3" type="audio/mp4"></audio>
718
- </td>
719
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/musicgen/techno_MDDBRockabilly_musicgen.mp3" type="audio/mp4"></audio></td>
720
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/ddim/techno_MDDBRockabilly_ddim.mp3" type="audio/mp4"></audio></td>
721
- </tr>
722
- <tr class="align-top" hidden>
723
- <th scope="row">7</th>
724
- <td class="align-top text-left">A recording of a <b>sneaky jazz song</b>.</td>
725
- <td class="align-top">A recording of a <b>tense classical music score</b>.</td>
726
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/orig/MDDBFreeJazz.mp3" type="audio/mp4"></audio></td>
727
- <td class="text-center align-middle"><br>
728
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sup_samples/orchestra_MDDBFreeJazz_ours_90.mp3" type="audio/mp4"></audio>
729
- <br>T<sub>start</sub>=110</td>
730
- <td class="text-center align-middle"><br>
731
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/orchestra_MDDBFreeJazz_sdedit_100.mp3" type="audio/mp4"></audio>
732
- <br>
733
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/orchestra_MDDBFreeJazz_sdedit_130.mp3" type="audio/mp4"></audio>
734
- <br>
735
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/orchestra_MDDBFreeJazz_sdedit_160.mp3" type="audio/mp4"></audio>
736
- </td>
737
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/musicgen/orchestra_MDDBFreeJazz_musicgen.mp3" type="audio/mp4"></audio></td>
738
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/ddim/orchestra_MDDBFreeJazz_ddim.mp3" type="audio/mp4"></audio></td>
739
- </tr>
740
- <tr class="align-top" hidden>
741
- <th scope="row">8</th>
742
- <td class="align-top text-center">—</td>
743
- <td class="align-top">A recording of a funky hip hop song.</td>
744
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/orig/MDDBFunkJazz.mp3" type="audio/mp4"></audio></td>
745
- <td class="text-center align-middle"><br>
746
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sup_samples/hiphop_MDDBFunkJazz_ours_90.mp3" type="audio/mp4"></audio>
747
- <br>T<sub>start</sub>=110</td>
748
- <td class="text-center align-middle"><br>
749
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/hiphop_MDDBFunkJazz_sdedit_100.mp3" type="audio/mp4"></audio>
750
- <br>
751
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/hiphop_MDDBFunkJazz_sdedit_130.mp3" type="audio/mp4"></audio>
752
- <br>
753
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/hiphop_MDDBFunkJazz_sdedit_160.mp3" type="audio/mp4"></audio>
754
- </td>
755
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/musicgen/hiphop_MDDBFunkJazz_musicgen.mp3" type="audio/mp4"></audio></td>
756
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/ddim/hiphop_MDDBFunkJazz_ddim.mp3" type="audio/mp4"></audio></td>
757
- </tr>
758
- <tr class="align-top" hidden>
759
- <th scope="row">9</th>
760
- <td class="align-top">A recording of an upbeat <b>gospel</b> song.</td>
761
- <td class="align-top">A recording of an upbeat <b>techno</b> song.</td>
762
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/orig/MDDBGospel.mp3" type="audio/mp4"></audio></td>
763
- <td class="text-center align-middle"><br>
764
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sup_samples/techno_MDDBGospel_ours_100.mp3" type="audio/mp4"></audio>
765
- <br>T<sub>start</sub>=100</td>
766
- <td class="text-center align-middle"><br>
767
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/techno_MDDBGospel_sdedit_100.mp3" type="audio/mp4"></audio>
768
- <br>
769
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/techno_MDDBGospel_sdedit_130.mp3" type="audio/mp4"></audio>
770
- <br>
771
- <audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/sdedit/techno_MDDBGospel_sdedit_160.mp3" type="audio/mp4"></audio>
772
- </td>
773
- <td class="text-center align-middle"><audio preload="auto" class="shorteraudio" controls><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/audio/musicgen/techno_MDDBGospel_musicgen.mp3" type="audio/mp4"></audio></td>
774
- <td class="text-center align-middle"><audio></audio>
 
1
  <!DOCTYPE html>
2
  <html lang="en">
3
  <head>
4
+ <meta charset="utf-8">
5
+ <meta name="viewport" content="width=device-width, initial-scale=1">
6
+ <title>Zero-Shot Unsupervised and Text-Based Audio Editing Using DDPM Inversion</title>
7
+ <link rel="shortcut icon" href="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/favicon.ico">
8
+ <meta property="og:image" content="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/teaser.png"/>
9
+ <meta property="og:title" content="Zero-Shot Unsupervised and Text-Based Audio Editing Using DDPM Inversion" />
10
+ <meta property="og:description" content="Zero-Shot Unsupervised and Text-Based Audio Editing Using DDPM Inversion" />
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
11
 
12
+ <!-- Tailwind CSS -->
13
+ <script src="https://cdn.tailwindcss.com"></script>
14
+ <!-- Google Fonts: Poppins -->
15
+ <link href="https://fonts.googleapis.com/css2?family=Poppins:wght@400;600;700&display=swap" rel="stylesheet">
16
+ <!-- Font Awesome -->
17
+ <script src="https://kit.fontawesome.com/ad96f96272.js" crossorigin="anonymous"></script>
18
+ <!-- Academicons -->
19
+ <link rel="stylesheet" href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
20
+
21
+ <style>
22
+ body {
23
+ font-family: 'Poppins', sans-serif;
24
+ background: linear-gradient(135deg, #1e293b 0%, #0f172a 100%);
25
+ color: #f1f5f9;
26
+ min-height: 100vh;
27
+ }
28
+ .glass {
29
+ background: rgba(30, 41, 59, 0.7);
30
+ box-shadow: 0 8px 32px 0 rgba(31, 38, 135, 0.37);
31
+ backdrop-filter: blur(8px);
32
+ border-radius: 1rem;
33
+ border: 1px solid rgba(255,255,255,0.08);
34
+ }
35
+ .gradient-btn {
36
+ background: linear-gradient(90deg, #2563eb 0%, #1e40af 100%);
37
+ color: #fff;
38
+ transition: transform 0.2s, box-shadow 0.2s;
39
+ }
40
+ .gradient-btn:hover {
41
+ transform: translateY(-2px) scale(1.03);
42
+ box-shadow: 0 4px 20px 0 rgba(37,99,235,0.25);
43
+ background: linear-gradient(90deg, #1e40af 0%, #2563eb 100%);
44
+ }
45
+ .audio-player {
46
+ width: 100%;
47
+ max-width: 260px;
48
+ background: #1e293b;
49
+ border-radius: 0.5rem;
50
+ box-shadow: 0 2px 8px rgba(16, 24, 39, 0.2);
51
+ padding: 0.5rem;
52
+ }
53
+ ::selection {
54
+ background: #2563eb;
55
+ color: #fff;
56
+ }
57
+ </style>
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
58
  </head>
59
  <body>
60
+ <!-- Navigation Bar -->
61
+ <nav class="glass sticky top-0 z-50 flex items-center justify-between px-6 py-4 shadow-lg">
62
+ <a href="#" class="text-2xl font-bold tracking-tight bg-gradient-to-r from-blue-400 to-indigo-500 bg-clip-text text-transparent">Zero-Shot Audio Editing</a>
63
+ <div class="hidden md:flex space-x-6">
64
+ <a href="#supsamples" class="hover:text-blue-400 transition">Text-based Editing</a>
65
+ <a href="#unsupsamples" class="hover:text-blue-400 transition">Unsup. Editing</a>
66
+ <a href="#supcomparisons" class="hover:text-blue-400 transition">Text-Based Editing Comp.</a>
67
+ <a href="#unsupcomparisons" class="hover:text-blue-400 transition">Unsup. Editing Comp.</a>
68
+ </div>
69
+ <button class="md:hidden text-2xl"><i class="fas fa-bars"></i></button>
70
+ </nav>
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
71
 
72
+ <!-- Header Section -->
73
+ <header class="max-w-3xl mx-auto mt-12 text-center glass py-10 px-6">
74
+ <h1 class="text-4xl md:text-5xl font-extrabold mb-4 bg-gradient-to-r from-blue-400 to-indigo-500 bg-clip-text text-transparent">
75
+ Zero-Shot Unsupervised and Text-Based Audio Editing Using DDPM Inversion
76
+ </h1>
77
+ <h4 class="text-xl md:text-2xl font-semibold mb-2">ICML 2024</h4>
78
+ <div class="flex flex-wrap justify-center gap-4 mb-2">
79
+ <a href="https://www.linkedin.com/in/hilamanor/" target="_blank" class="text-lg hover:text-blue-400 transition">Hila Manor</a>
80
+ <span class="text-gray-400">|</span>
81
+ <a href="https://tomer.net.technion.ac.il/" target="_blank" class="text-lg hover:text-blue-400 transition">Tomer Michaeli</a>
82
+ </div>
83
+ <p class="mb-4 text-gray-300">Technion - Israel Institute of Technology</p>
84
+ <div class="flex flex-wrap justify-center gap-4 mt-4">
85
+ <a class="gradient-btn px-5 py-2 rounded-lg font-semibold flex items-center gap-2" href="https://arxiv.org/abs/2402.10009" target="_blank"><i class="ai ai-arxiv"></i>ArXiv</a>
86
+ <a class="gradient-btn px-5 py-2 rounded-lg font-semibold flex items-center gap-2" href="https://github.com/HilaManor/AudioEditingCode/" target="_blank"><i class="fab fa-github"></i>Code</a>
87
+ <a class="gradient-btn px-5 py-2 rounded-lg font-semibold flex items-center gap-2" href="https://youtu.be/lBnldOQVbS4" target="_blank"><i class="fab fa-youtube"></i>Presentation</a>
88
+ <a class="gradient-btn px-5 py-2 rounded-lg font-semibold flex items-center gap-2" href="https://huggingface.co/spaces/hilamanor/audioEditing/" target="_blank">🤗 Text-Based Space</a>
89
+ </div>
90
+ </header>
91
 
92
+ <!-- Abstract and Teaser -->
93
+ <section class="max-w-3xl mx-auto my-12 glass py-8 px-6 text-center">
94
+ <img alt="Teaser" src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/teaser.png" class="mx-auto rounded-xl shadow-xl mb-6 w-full md:w-3/4"/>
95
+ <h2 class="text-2xl font-semibold mb-4">Abstract</h2>
96
+ <p class="text-lg leading-relaxed text-gray-200">
97
+ Editing signals using large pre-trained models, in a zero-shot manner, has recently seen rapid advancements in the image domain. However, this wave has yet to reach the audio domain.
98
+ In this paper, we explore two zero-shot editing techniques for audio signals, which use DDPM inversion on pre-trained diffusion models. The first, adopted from the image domain, allows text-based editing.
99
+ The second, is a novel approach for discovering semantically meaningful editing directions without supervision.
100
+ When applied to music signals, this method exposes a range of musically interesting modifications, from controlling the participation of specific instruments to improvisations on the melody.
101
+ </p>
102
+ </section>
103
 
104
+ <!-- Video Overview -->
105
+ <section class="max-w-3xl mx-auto my-12 glass py-8 px-6 text-center">
106
+ <h2 class="text-2xl font-semibold mb-2">Video Overview</h2>
107
+ <p class="text-gray-300 mb-4">For people in a hurry. Images generated by <a href="https://openai.com/dall-e-2" class="text-blue-400 hover:underline">DALL-E 2</a> and <a href="https://www.bing.com/copilot" class="text-blue-400 hover:underline">Copilot</a>.</p>
108
+ <video controls class="mx-auto rounded-xl shadow-xl w-full max-w-md">
109
+ <source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/overview.mp4" type="video/mp4">
110
+ Your browser does not support the video tag.
111
+ </video>
112
+ </section>
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
113
 
114
+ <!-- Samples of Editing -->
115
+ <section class="max-w-5xl mx-auto my-12 glass py-8 px-6">
116
+ <h2 class="text-2xl font-semibold text-center mb-4">1. Samples of Editing</h2>
117
+ <p class="text-center text-gray-300 mb-8">We present samples of audio editing using our proposed methods. The samples are organized into two sections: text-based editing and unsupervised editing.</p>
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
118
 
119
+ <!-- Text-Based Editing Table -->
120
+ <div>
121
+ <h3 id="supsamples" class="text-xl font-semibold mb-4">1.1. Samples of Text-Based Editing</h3>
122
+ <div class="overflow-x-auto">
123
+ <table class="min-w-full text-sm text-left text-gray-200">
124
+ <thead class="bg-blue-900/80">
125
+ <tr>
126
+ <th class="px-4 py-2">#</th>
127
+ <th class="px-4 py-2">Source Prompt</th>
128
+ <th class="px-4 py-2">Target Prompt</th>
129
+ <th class="px-4 py-2">Original Audio</th>
130
+ <th class="px-4 py-2">Edited Audio</th>
131
+ <th class="px-4 py-2">Edit T<sub>start</sub></th>
132
+ </tr>
133
+ </thead>
134
+ <tbody>
135
+ <tr class="hover:bg-blue-800/30 transition">
136
+ <td class="px-4 py-2">1</td>
137
+ <td class="px-4 py-2">A recording of a <b>sneaky jazz</b> song.</td>
138
+ <td class="px-4 py-2">A recording of a <b>tense classical</b> music score.</td>
139
+ <td class="px-4 py-2"><audio controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/orig/MDDBFreeJazz.mp3" type="audio/mp4"></audio></td>
140
+ <td class="px-4 py-2"><audio controls class="audio-player"><source src="https://github.com/HilaManor/AudioEditingCode/raw/refs/heads/codeclean/docs/resources/audio/sup_samples/orchestra_MDDBFreeJazz_ours_90.mp3" type="audio/mp4"></audio></td>
141
+ <td class="px-4 py-2 text-center">110</td>
142
+ </tr>
143
+ <!-- Repeat for other rows as needed ... -->
144
+ </tbody>
145
+ </table>
146
+ </div>
147
+ </div>
148
+ <!-- Add more sections as needed -->
149
+ </section>
150
 
151
+ <!-- Back to Top Button -->
152
+ <button onclick="window.scrollTo({top: 0, behavior: 'smooth'});" class="fixed bottom-8 right-8 gradient-btn rounded-full p-4 shadow-xl hover:scale-110 transition hidden md:block" aria-label="Back to top">
153
+ <i class="fas fa-arrow-up"></i>
154
+ </button>
155
 
156
+ <script>
157
+ // Show back-to-top button on scroll
158
+ window.addEventListener('scroll', function() {
159
+ const btn = document.querySelector('.fixed.bottom-8.right-8');
160
+ if (window.scrollY > 300) {
161
+ btn.classList.remove('hidden');
162
+ } else {
163
+ btn.classList.add('hidden');
164
+ }
165
+ });
166
+ </script>
167
+ </body>
168
+ </html>