icassp 2025

Author: gabolsgabs

_posts/2025-04-06-ICASSP-Afchar.md

@@ -0,0 +1,23 @@
+---
+layout: post
+title: "AI-Generated Music Detection and its Challenges"
+date: 2025-04-10 10:00:00 +0200
+category: Publication
+author: dafchar
+readtime: 1
+domains: 
+ - MIR
+people:
+ - dafchar
+ - gmeseguerbrocal
+ - rhennequin
+publication_type: conference
+publication_title: "AI-Generated Music Detection and its Challenges"
+publication_year: 2025
+publication_authors: Darius Afchar, Gabriel Meseguer-Brocal, Romain Hennequin
+publication_conference: ICASSP
+publication_preprint: "https://arxiv.org/pdf/2501.10111"
+publication_code: "https://github.com/deezer/deepfake-detector"
+---
+
+In the face of a new era of generative models, the detection of artificially generated content has become a matter of utmost importance. In particular, the ability to create credible minute-long synthetic music in a few seconds on user-friendly platforms poses a real threat of fraud on streaming services and unfair competition to human artists. This paper demonstrates the possibility (and surprising ease) of training classifiers on datasets comprising real audio and artificial reconstructions, achieving a convincing accuracy of 99.8%. To our knowledge, this marks the first publication of a AI-music detector, a tool that will help in the regulation of synthetic media. Nevertheless, informed by decades of literature on forgery detection in other fields, we stress that getting a good test score is not the end of the story. We expose and discuss several facets that could be problematic with such a deployed detector: robustness to audio manipulation, generalisation to unseen models. This second part acts as a position for future research steps in the field and a caveat to a flourishing market of artificial content checkers.

_posts/2025-04-06-ICASSP-Kong.md

@@ -0,0 +1,23 @@
+---
+layout: post
+title: "S-KEY: Self-supervised Learning of Major and Minor Keys from Audio"
+date: 2025-04-10 10:00:00 +0200
+category: Publication
+author: ykong
+readtime: 1
+domains: 
+ - MIR
+people:
+ - ykong
+ - gmeseguerbrocal
+ - rhennequin
+publication_type: conference
+publication_title: "S-KEY: Self-supervised Learning of Major and Minor Keys from Audio"
+publication_year: 2025
+publication_authors: Yuexuan Kong, Gabriel Meseguer-Brocal, Vincent Lostanlen, Mathieu Lagrange, Romain Hennequin
+publication_conference: ICASSP
+publication_preprint: "https://arxiv.org/pdf/2501.12907"
+publication_code: "https://github.com/deezer/s-key"
+---
+
+STONE, the current method in self-supervised learning for tonality estimation in music signals, cannot distinguish relative keys, such as C major versus A minor. In this article, we extend the neural network architecture and learning objective of STONE to perform self-supervised learning of major and minor keys (S-KEY). Our main contribution is an auxiliary pretext task to STONE, formulated using transposition-invariant chroma features as a source of pseudo-labels. S-KEY matches the supervised state of the art in tonality estimation on FMAKv2 and GTZAN datasets while requiring no human annotation and having the same parameter budget as STONE. We build upon this result and expand the training set of S-KEY to a million songs, thus showing the potential of large-scale self-supervised learning in music information retrieval.