Paper published in a journal (Scientific congresses, symposiums and conference proceedings)
Multi-label Deepfake Classification
SINGH, Inder Pal; MEJRI, Nesryne; NGUYEN, van Dat et al.
2023In IEEE Workshop on Multimedia Signal Processing
Peer reviewed
 

Files


Full Text
DeepFake_MLIC_InderPalSINGH_MMSP2023_camera_ready .pdf
Author postprint (279.52 kB)
Download

All documents in ORBilu are protected by a user license.

Send to



Details



Keywords :
Deepfakes; Multi-label Image Classification; Deep Learning; Machine Learning
Abstract :
[en] In this paper, we investigate the suitability of current multi-label classification approaches for deepfake detection. With the recent advances in generative modeling, new deepfake detection methods have been proposed. Nevertheless, they mostly formulate this topic as a binary classification problem, resulting in poor explainability capabilities. Indeed, a forged image might be induced by multi-step manipulations with different properties. For a better interpretability of the results, recognizing the nature of these stacked manipulations is highly relevant. For that reason, we propose to model deepfake detection as a multi-label classification task, where each label corresponds to a specific kind of manipulation. In this context, state-of-the-art multi-label image classification methods are considered. Extensive experiments are performed to assess the practical use case of deepfake detection.
Research center :
SnT - Interdisciplinary Centre for Security, Reliability and Trust, University of Luxembourg
Disciplines :
Computer science
Author, co-author :
SINGH, Inder Pal  ;  University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > CVI2
MEJRI, Nesryne  ;  University of Luxembourg > Faculty of Science, Technology and Medicine (FSTM) > CVI2
NGUYEN, van Dat ;  University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > CVI2
Ghorbel, Enjie
AOUADA, Djamila  ;  University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > CVI2
External co-authors :
no
Language :
English
Title :
Multi-label Deepfake Classification
Publication date :
27 September 2023
Event name :
The IEEE 25th International Workshop on Multimedia Workshop
Event organizer :
Multimedia Signal Processing Technical Committee of the IEEE Signal Processing Society (SPS)
Event place :
Poitiers, France
Event date :
from 27-09-2023 to 29-09-2023
Audience :
International
Journal title :
IEEE Workshop on Multimedia Signal Processing
Peer reviewed :
Peer reviewed
Focus Area :
Security, Reliability and Trust
FnR Project :
FNR16353350 - Deepfake Detection Using Spatio-temporal-spectral Representations For Effective Learning, 2021 (01/03/2022-28/02/2025) - Djamila Aouada
Funders :
FNR - Fonds National de la Recherche [LU]
Commentary :
This work is supported by the Luxembourg National Research Fund, under the BRIDGES2021/IS/16353350/FaKeDeTeR and UNFAKE, ref.16763798 projects, and by Post Luxembourg.
Available on ORBilu :
since 10 September 2023

Statistics


Number of views
256 (19 by Unilu)
Number of downloads
140 (29 by Unilu)

Scopus citations®
 
1
Scopus citations®
without self-citations
0

Bibliography


Similar publications



Contact ORBilu