-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathpublications.html
327 lines (246 loc) · 21.3 KB
/
publications.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
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
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
<html lang="fr"><head>
<meta charset="utf-8">
<head>
<title>Publications</title>
<link rel="stylesheet" href="/assets/css/global.css">
<link rel="stylesheet" href="/assets/css/profile.css">
<link rel="stylesheet" href="/assets/css/index.css">
<link rel="stylesheet" href="/assets/css/box_links.css">
<script>
function open_or_close(id){
var myDiv = document.getElementById(id)
if (myDiv.style.display=='block'){
myDiv.style.display='none';
} else {
myDiv.style.display='block';
}
};
</script>
<script src="https://kit.fontawesome.com/12b596ef39.js" crossorigin="anonymous"></script>
<link type="application/atom+xml" rel="alternate" href="/feed.xml" />
<!-- Begin Jekyll SEO tag v2.8.0 -->
<title>Publications</title>
<meta name="generator" content="Jekyll v4.3.2" />
<meta property="og:title" content="Publications" />
<meta property="og:locale" content="en_US" />
<meta property="og:type" content="website" />
<meta name="twitter:card" content="summary" />
<meta property="twitter:title" content="Publications" />
<script type="application/ld+json">
{"@context":"https://schema.org","@type":"WebPage","headline":"Publications","url":"/publications"}</script>
<!-- End Jekyll SEO tag -->
</head>
<body>
<div class="content">
<div class="base">
<div class="profile">
<div class="photo">
<img src="/assets/images/light_photo_sympa.jpeg"/>
</div>
<div class="info">
<h1 class="name"><a href="/">Axel Marmoret</a></h1>
<h2 class="job">Assistant Professor at IMT Atlantique (BRAIn team)</h2>
<h4>PhD Graduate in Signal Processing, Computer Science Engineer</h4>
</div>
</div>
<div class="about">
<h3>About Me</h3>
Hello! My name is Axel Marmoret, I'm 26, and I'm an Assistant Professor at IMT Atlantique in France.<br>
I work on Machine Learning and Optimisation techniques for rerieving musical structure, based on its audio form.<br>
I'm also a musician (mainly drummer, newly bassist!), and passionate about music.
</div>
<div class="contact">
<h3>Contact Me</h3>
<div class="email"><a href="mailto:[email protected]"><i class="fas fa-envelope"></i><span>My mail adress</span></a></div>
<div class="address"><a href="https://goo.gl/maps/K8BJVp8kKWW9CQZt6" target="_blank"><i class="fas fa-map-marker"></i><span>Living in Brest, France</span></a></div>
</div>
<div class="follow">
<h3>Follow Me</h3>
<div class="box">
<a href="https://gitlab.inria.fr/amarmore" target="_blank"><i class="fab fa-gitlab "></i></a>
<a href="https://hal.archives-ouvertes.fr/search/index/?q=%2A&authFullName_s=Axel+Marmoret" target="_blank"><i class="far fa-file-alt"></i></a>
<a href="https://www.linkedin.com/in/axel-marmoret-phd-%E2%8F%9A-732491136/" target="_blank"><i class="fab fa-linkedin"></i></a>
<a href="https://open.spotify.com/user/e-xa" target="_blank"><i class="fab fa-spotify"></i></a>
</div>
</div>
</div>
<div class="func">
<ul id="navigation">
<li><a href="/research/">Research</a></li>
<li><a href="/resume/">Resume</a></li>
<li><a href="/publications/">Publications</a></li>
<li><a href="/codes/">Codes</a></li>
<li><a href="/TODO/">Music</a></li>
<li><a href="/resources/">Resources</a></li>
</ul>
<h1>International Conferences</h1>
<h2>Accepted</h2>
<div class="content_block">
<h2>Convolutive Block-Matching Segmentation Algorithm with Application to Music Structure Analysis.</h2>
Axel Marmoret, Jérémy Cohen, Frédéric Bimbot. IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, Mohonk Mountain House, New Paltz, NY, USA, Oct 22-25, 2023.
<br><br>
<div class="box">
<a href="javascript:void(0);" onclick="open_or_close('waspaa_2023_abstract')"><i class="fas fa-info-circle"></i></a>
<a href="javascript:void(0);" onclick="open_or_close('waspaa_2023_bibtex')"><i class="fas fa-quote-left"></i></a>
<a href=https://hal.archives-ouvertes.fr/hal-03834996v1 target="_blank"><i class="fas fa-unlock-alt"></i></a>
<a href=https://gitlab.inria.fr/amarmore/autosimilarity_segmentation target="_blank"><i class="fab fa-gitlab"></i></a>
</div>
<div id="waspaa_2023_bibtex" class="hidden_content_block">
<div class="citation_content">
<p>@inproceedings{marmoret2023convolutive, title={Convolutive Block-Matching Segmentation Algorithm with Application to Music Structure Analysis}, author={Marmoret, Axel and Cohen, J{\'e}r{\'e}my E and Bimbot, Fr{\'e}d{\'e}ric}, booktitle={WASPAA}, year={2023}}</p>
</div>
</div>
<div id="waspaa_2023_abstract" class="hidden_content_block">
<div class="citation_content">
<p>Abstract: Music Structure Analysis (MSA) consists of representing a song in sections (such as 'chorus', 'verse', 'solo' etc), and can be seen as the retrieval of a simplified organization of the song. This work presents a new algorithm, called Convolutive Block-Matching (CBM) algorithm, devoted to MSA. In particular, the CBM algorithm is a dynamic programming algorithm, applying on autosimilarity matrices, a standard tool in MSA. In this work, autosimilarity matrices are computed from the feature representation of an audio signal, and time is sampled on the barscale. We study three different similarity functions for the computation of autosimilarity matrices. We report that the proposed algorithm achieves a level of performance competitive to that of supervised state-of-the-art methods on 3 among 4 metrics, while being fully unsupervised.</p>
</div>
</div>
</div>
<div class="content_block">
<h2>Barwise Compression Schemes for Audio-Based Music Structure Analysis.</h2>
Axel Marmoret, Jérémy Cohen, Frédéric Bimbot. SMC 2022 - 19th Sound and Music Computing Conference, Jun 2022, Saint-Etienne, France.
<br><br>
<div class="box">
<a href="javascript:void(0);" onclick="open_or_close('smc_2022_barcomp_abstract')"><i class="fas fa-info-circle"></i></a>
<a href="javascript:void(0);" onclick="open_or_close('smc_2022_barcomp_bibtex')"><i class="fas fa-quote-left"></i></a>
<a href=https://hal.science/hal-03600873 target="_blank"><i class="fas fa-unlock-alt"></i></a>
<a href=https://gitlab.inria.fr/amarmore/barwisemusiccompression target="_blank"><i class="fab fa-gitlab"></i></a>
<a href=/assets/pdf/posters/SMC22_barwise.pdf target="_blank"><i class="fab fa-slideshare"></i></a>
</div>
<div id="smc_2022_barcomp_bibtex" class="hidden_content_block">
<div class="citation_content">
<p>@inproceedings{Barwise Compression Schemes for Audio-Based Music Structure Analysis}, author={Marmoret, Axel and Cohen, J{\'e}r{\'e}my E and Bimbot, Fr{\'e}d{\'e}ric}, booktitle={{Proceedings of the 19th Sound and Music Computing Conference}}, year={2022}, doi= {10.5281/zenodo.6822204}}</p>
</div>
</div>
<div id="smc_2022_barcomp_abstract" class="hidden_content_block">
<div class="citation_content">
<p>Abstract: Music Structure Analysis (MSA) consists in segmenting a music piece in several distinct sections. We approach MSA within a compression framework, under the hypothesis that the structure is more easily revealed by a simplified representation of the original content of the song. More specifically, under the hypothesis that MSA is correlated with similarities occurring at the bar scale, this article introduces the use of linear and non-linear compression schemes on barwise audio signals. Compressed representations capture the most salient components of the different bars in the song and are then used to infer the song structure using a dynamic programming algorithm. This work explores both low-rank approximation models such as Principal Component Analysis or Nonnegative Matrix Factorization and ``piece-specific'' Auto-Encoding Neural Networks, with the objective to learn latent representations specific to a given song. Such approaches do not rely on supervision nor annotations, which are well-known to be tedious to collect and possibly ambiguous in MSA description. In our experiments, several unsupervised compression schemes achieve a level of performance comparable to that of state-of-the-art supervised methods (for 3s tolerance) on the RWC-Pop dataset, showcasing the importance of the barwise compression processing for MSA.</p>
</div>
</div>
</div>
<div class="content_block">
<h2>Semi-Supervised Convolutive NMF for Automatic Piano Transcription.</h2>
Haoran Wu, Axel Marmoret, Jérémy Cohen. SMC 2022 - 19th Sound and Music Computing Conference, Jun 2022, Saint-Etienne, France.
<br><br>
<div class="box">
<a href="javascript:void(0);" onclick="open_or_close('smc_2022_cnmf_abstract')"><i class="fas fa-info-circle"></i></a>
<a href="javascript:void(0);" onclick="open_or_close('smc_2022_cnmf_bibtex')"><i class="fas fa-quote-left"></i></a>
<a href=https://hal.science/hal-03608497 target="_blank"><i class="fas fa-unlock-alt"></i></a>
<a href=https://github.com/cohenjer/TransSSCNMF target="_blank"><i class="fab fa-github"></i></a>
<a href=/assets/pdf/posters/SMC22_semi.pdf target="_blank"><i class="fab fa-slideshare"></i></a>
</div>
<div id="smc_2022_cnmf_bibtex" class="hidden_content_block">
<div class="citation_content">
<p>@inproceedings{wu2022semi, title={Semi-Supervised Convolutive NMF for Automatic Music Transcription}, author={Wu, Haoran and Marmoret, Axel and Cohen, J{\'e}r{\'e}my E}, booktitle={{Proceedings of the 19th Sound and Music Computing Conference}}, year={2022}, doi= {10.5281/zenodo.6822204}}</p>
</div>
</div>
<div id="smc_2022_cnmf_abstract" class="hidden_content_block">
<div class="citation_content">
<p>Abstract: Automatic Music Transcription, which consists in transforming an audio recording of a musical performance into symbolic format, remains a difficult Music Information Retrieval task. In this work, which focuses on piano transcription, we propose a semi-supervised approach using low-rank matrix factorization techniques, in particular Convolutive Nonnegative Matrix Factorization. In the semi-supervised setting, only a single recording of each individual notes is required. We show on the MAPS dataset that the proposed semi-supervised CNMF method performs better than state-of-the-art low-rank factorization techniques and a little worse than supervised deep learning state-of-the-art methods, while however suffering from generalization issues.</p>
</div>
</div>
</div>
<div class="content_block">
<h2>Uncovering Audio Patterns in Music with Nonnegative Tucker Decomposition for Structural Segmentation.</h2>
Axel Marmoret, Jérémy Cohen, Nancy Bertin, Frédéric Bimbot. ISMIR 2020 - 21st International Society for Music Information Retrieval, Oct 2020, Montréal (Online), Canada.
<br><br>
<div class="box">
<a href="javascript:void(0);" onclick="open_or_close('ismir_2020_abstract')"><i class="fas fa-info-circle"></i></a>
<a href="javascript:void(0);" onclick="open_or_close('ismir_2020_bibtex')"><i class="fas fa-quote-left"></i></a>
<a href=https://hal.science/hal-02928733v1 target="_blank"><i class="fas fa-unlock-alt"></i></a>
<a href=https://gitlab.inria.fr/amarmore/musicntd/-/tree/0.1.0 target="_blank"><i class="fab fa-gitlab"></i></a>
<a href=https://youtu.be/qN8_y-k9KxI target="_blank"><i class="fab fa-youtube"></i></a>
<a href=/assets/pdf/posters/ISMIR20_uncovering.pdf target="_blank"><i class="fab fa-slideshare"></i></a>
</div>
<div id="ismir_2020_bibtex" class="hidden_content_block">
<div class="citation_content">
<p>@inproceedings{marmoret2020uncovering, title={Uncovering Audio Patterns in Music with Nonnegative Tucker Decomposition for Structural Segmentation}, author={Marmoret, Axel and Cohen, J{\'e}r{\'e}my and Bertin, Nancy and Bimbot, Fr{\'e}d{\'e}ric}, booktitle={ISMIR 2020-21st International Society for Music Information Retrieval}, year={2020}}</p>
</div>
</div>
<div id="ismir_2020_abstract" class="hidden_content_block">
<div class="citation_content">
<p>Abstract: Recent work has proposed the use of tensor decomposition to model repetitions and to separate tracks in loop-based electronic music. The present work investigates further on the ability of Nonnegative Tucker Decompositon (NTD) to uncover musical patterns and structure in pop songs in their audio form. Exploiting the fact that NTD tends to express the content of bars as linear combinations of a few patterns, we illustrate the ability of the decomposition to capture and single out repeated motifs in the corresponding compressed space, which can be interpreted from a musical viewpoint. The resulting features also turn out to be efficient for structural segmentation, leading to experimental results on the RWC Pop data set which are potentially challenging state-of-the-art approaches that rely on extensive example-based learning schemes.</p>
</div>
</div>
</div>
<h1>National Conferences</h1>
<h2>Accepted</h2>
<div class="content_block">
<h2>Nonnegative Tucker Decomposition with Beta-divergence for Music Structure Analysis of audio signals.</h2>
Axel Marmoret, Florian Voorwinden, Valentin Leplat, Jérémy E Cohen, Frédéric Bimbot. GRETSI 2022: XXVIIIe Colloque, Sep 2022, Nancy, France.
<br><br>
<div class="box">
<a href="javascript:void(0);" onclick="open_or_close('gretsi_2022_abstract')"><i class="fas fa-info-circle"></i></a>
<a href="javascript:void(0);" onclick="open_or_close('gretsi_2022_bibtex')"><i class="fas fa-quote-left"></i></a>
<a href=https://hal.science/hal-03409508v1 target="_blank"><i class="fas fa-unlock-alt"></i></a>
<a href=https://gitlab.inria.fr/amarmore/nonnegative-factorization/-/tree/v0.2.0 target="_blank"><i class="fab fa-gitlab"></i></a>
<a href=/assets/pdf/posters/GRETSI22_nonnegative.pdf target="_blank"><i class="fab fa-slideshare"></i></a>
</div>
<div id="gretsi_2022_bibtex" class="hidden_content_block">
<div class="citation_content">
<p>@inproceedings{marmoret2022nonnegative, title={Nonnegative Tucker Decomposition with Beta-divergence for Music Structure Analysis of audio signals}, author={Marmoret, Axel and Voorwinden, Florian and Leplat, Valentin and Cohen, J{\'e}r{\'e}my E and Bimbot, Fr{\'e}d{\'e}ric}, booktitle={{GRETSI 2022: XXVIIIe Colloque}}, year={2022}, organization={GRETSI}}</p>
</div>
</div>
<div id="gretsi_2022_abstract" class="hidden_content_block">
<div class="citation_content">
<p>Abstract: Nonnegative Tucker Decomposition (NTD), a tensor decomposition model, has received increased interest in the recent years because of its ability to blindly extract meaningful patterns in tensor data. Nevertheless, existing algorithms to compute NTD are mostly designed for the Euclidean loss. On the other hand, NTD has recently proven to be a powerful tool in Music Information Retrieval. This work proposes a Multiplicative Updates algorithm to compute NTD with the beta-divergence loss, often considered a better loss for audio processing. We notably show how to implement efficiently the multiplicative rules using tensor algebra, a naive approach being intractable. Finally, we show on a Music Structure Analysis task that unsupervised NTD fitted with beta-divergence loss outperforms earlier results obtained with the Euclidean loss.</p>
</div>
</div>
</div>
<h1>Thesis</h1>
<div class="content_block">
<h2>Unsupervised Machine Learning Paradigms for the Representation of Music Similarity and Structure.</h2>
Axel Marmoret. PhD thesis in Signal Processing, Université Rennes 1. Under the supervision of Jérémy Cohen and Frédéric Bimbot, with the additional help of Nancy Bertin and Simon Leglaive.
<br><br>
<div class="box">
<a href="javascript:void(0);" onclick="open_or_close('phd_abstract')"><i class="fas fa-info-circle"></i></a>
<a href="javascript:void(0);" onclick="open_or_close('phd_bibtex')"><i class="fas fa-quote-left"></i></a>
<a href=https://hal.science/tel-03937846 target="_blank"><i class="fas fa-unlock-alt"></i></a>
<a href=https://www.youtube.com/watch?v=Mw7WiuDlQjs target="_blank"><i class="fab fa-youtube"></i></a>
<a href=https://gitlab.inria.fr/amarmore/autosimilarity_segmentation target="_blank"><i class="fab fa-gitlab"></i></a>
<a href=https://gitlab.inria.fr/amarmore/nonnegative-factorization target="_blank"><i class="fab fa-gitlab"></i></a>
<a href=https://gitlab.inria.fr/amarmore/barmuscomp target="_blank"><i class="fab fa-gitlab"></i></a>
</div>
<div id="phd_bibtex" class="hidden_content_block">
<div class="citation_content">
<p>@phdthesis{marmoret2022unsupervised, title={Unsupervised Machine Learning Paradigms for the Representation of Music Similarity and Structure.}, author={Marmoret, Axel}, year={2022}, school={Universit{\'e} Rennes 1}}</p>
</div>
</div>
<div id="phd_abstract" class="hidden_content_block">
<div class="citation_content">
<p>Abstract: Musical structure, defined as a simplified representation of the organization of a song, is an important musicological concept, but hard to automatically estimate. This thesis presents new methods to automatically estimate the structural segmentation of a song, focusing the study of music at the barscale. By developing a new segmentation algorithm (called ``CBM'') and by comparing several unsupervised compression schemes (from linear and multilinear algebra to neural networks), paradigms introduced in this thesis result in segmentation performance outperforming those of the unsupervised State-of-the-Art methods and almost similar with those of the global State-of-the-Art, obtained with supervised machine learning algorithms. In particular, as the methods described in this thesis are unsupervised, the estimation do not rely on annotated data, lowering the bias in the estimates related to ambiguity and subjectivity (inherent to musical structure) while limiting the loss in performance compared to the best supervised methods. In addition, some of the methods studied in this thesis (in particular Nonnegative Tucker Decomposition) allow to extract automatically interpretable parts of a song which may be used for other task than the estimation of structure, and participate in the development of interpretable machine and deep learning algorithms, which is a major field of research nowadays.</p>
</div>
</div>
</div>
<div class="content_block">
<h2>Multi-Channel Automatic Music Transcription Using Tensor Algebra.</h2>
Axel Marmoret. Master's thesis. Under the supervision of Nancy Bertin and Jérémy Cohen. arXiv preprint arXiv:2107.11250.
<br><br>
<div class="box">
<a href="javascript:void(0);" onclick="open_or_close('master_abstract')"><i class="fas fa-info-circle"></i></a>
<a href="javascript:void(0);" onclick="open_or_close('master_bibtex')"><i class="fas fa-quote-left"></i></a>
<a href=https://hal.archives-ouvertes.fr/hal-03301448 target="_blank"><i class="fas fa-unlock-alt"></i></a>
<a href=https://gitlab.inria.fr/amarmore/nonnegative-factorization/ target="_blank"><i class="fab fa-gitlab"></i></a>
</div>
<div id="master_bibtex" class="hidden_content_block">
<div class="citation_content">
<p>@article{marmoret2019multi, title={Multi-Channel Automatic Music Transcription Using Tensor Algebra}, author={Marmoret, Axel and Bertin, Nancy and Cohen, J{\'e}r{\'e}my}, journal={arXiv preprint arXiv:2107.11250}, year={2019}}</p>
</div>
</div>
<div id="master_abstract" class="hidden_content_block">
<div class="citation_content">
<p>Abstract: Music is an art, perceived in unique ways by every listener, coming from acoustic signals. In the meantime, standards as musical scores exist to describe it. Even if humans can make this transcription, it is costly in terms of time and efforts, even more with the explosion of information consecutively to the rise of the Internet. In that sense, researches are driven in the direction of Automatic Music Transcription. While this task is considered solved in the case of single notes, it is still open when notes superpose themselves, forming chords. This report aims at developing some of the existing techniques towards Music Transcription, particularly matrix factorization, and introducing the concept of multi-channel automatic music transcription. This concept will be explored with mathematical objects called tensors.</p>
</div>
</div>
</div>
<div>
</div>
<footer>
<div class="copyright">
<p>For the theme: Copyright (c) 2020 by Naomi Bastian Weatherford <a target="_blank" class="inner-link" href="https://codepen.io/astronaomical/pen/KexYgb">(https://codepen.io/astronaomical/pen/KexYgb)</a></p>
</div>
</footer>
</div>
</div>
</body>
</html>