CAU KU deep fake detection system for ADD 2023 challenge*

Soyul Han, Taein Kang, Sunmook Choi, Jaejin Seo, Sanghyeok Chung, Sumi Lee, Seungsang Oh, Il Youp Kwak

Research output: Contribution to journalConference articlepeer-review

Abstract

The paper presents the participation of the CAU_KU team in the ADD 2023 Challenge, specifically in track 1.2 (audio fake game - detection track) and track 3 (deepfake algorithm recognition track). Various deep learning models were explored using features from the pretrained wav2vec2 network, as well as CQT, mel-spectrogram, etc. We modified the representation extraction component of the AASIST model to incorporate 2D spectrograms (wav2vec2 or CQT) and attempted different deep learning models, with model ensembling employed to create the final model. For track 1.2, our submitted ensemble model for round 1 utilized the CQT-LCNN and CQT-AASIST models. For round 2, our model used the CQT-LCNN, CQT-AASIST, and W2V2-GMM models. For track 3, we ensembled the CQT-LCNN, CQT-OFD and AASIST models. Additionally, we applied the openmax algorithm to detect unknown deepfake attacks. Our best submission achieved 23.44% and 21.26% on round 1 and 2 of track 1.2, respectively, and ranked 3rd in track 1.2.

Original languageEnglish
Pages (from-to)23-30
Number of pages8
JournalCEUR Workshop Proceedings
Volume3597
Publication statusPublished - 2023
Event2023 Workshop on Deepfake Audio Detection and Analysis, DADA 2023 - Macao, China
Duration: 2023 Aug 19 → …

Bibliographical note

Publisher Copyright:
© 2023 CEUR-WS. All rights reserved.

Keywords

  • audio deep synthesis
  • audio deepfake detection
  • deep learning
  • deepfake algorithm recognition

ASJC Scopus subject areas

  • General Computer Science

Fingerprint

Dive into the research topics of 'CAU KU deep fake detection system for ADD 2023 challenge*'. Together they form a unique fingerprint.

Cite this