Skip to main navigation Skip to search Skip to main content

Multilingual Music Deepfake Detection Using WavLM and MERT Audio Embeddings

  • Tejaswini Ponnada*
  • , Pawani Agarwal*
  • , G. Ignisha Rajathi
  • , R. Yasir Abdullah
  • , J. Mohanalin
  • , R. Johny Elton
  • *Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

With the growth in AI-generated material, a new issue has emerged in the form of deepfake music. These deepfakes are constructed with advanced models that can mimic human sounds or music in a variety of languages and accents. In this paper, we investigate the identification of multilingual music deepfakes - leveraging data from four languages: German, English, Spanish, and French. Regardless of the language or accent used, we aim to develop a system that can recognize phony music clips. We propose a novel framework that combines embeddings from WavLM-base-plus-sv, a speech and speaker-centric model, and MERT-v1-330M, an acoustic music understanding transformer, with a lightweight multilayer perceptron classifier to distinguish between real and deepfake music samples.

Original languageEnglish
Title of host publicationSoft Computing and Its Engineering Applications - 7th International Conference, icSoftComp 2025, Proceedings
EditorsKanubhai K. Patel, Atul Patel, KC Santosh, Gabriel Gomes de Oliveira, Ashish Ghosh
PublisherSpringer Science and Business Media Deutschland GmbH
Pages328-340
Number of pages13
ISBN (Print)9783032220585
DOIs
Publication statusPublished - 2026
Event7th International Conference on Soft Computing and its Engineering Applications, icSoftComp 2025 - Hanoi, Viet Nam
Duration: 09-12-202511-12-2025

Publication series

NameCommunications in Computer and Information Science
Volume2873 CCIS
ISSN (Print)1865-0929
ISSN (Electronic)1865-0937

Conference

Conference7th International Conference on Soft Computing and its Engineering Applications, icSoftComp 2025
Country/TerritoryViet Nam
CityHanoi
Period09-12-2511-12-25

All Science Journal Classification (ASJC) codes

  • General Computer Science
  • General Mathematics

Fingerprint

Dive into the research topics of 'Multilingual Music Deepfake Detection Using WavLM and MERT Audio Embeddings'. Together they form a unique fingerprint.

Cite this