arrow
返回

AVoiD-DF: Audio-Visual Joint Learning for Detecting Deepfake

delete2023-01-01
delete48
PRE
AI
W
Wenyuan Yang
X
Xiaoyu Zhou
Z
Zhikai Chen
B
Bofei Guo
Z
Zhongjie Ba
夏
夏志华 (Zhihua Xia) *
操晓春 封面图
操晓春 (Xiaochun Cao)
K
Kui Ren
DOI:10.1109/TIFS.2023.3262148delete
delete原文链接
delete原文求助
delete分享
delete收藏
摘要

摘要

En 中文
Recently, deepfakes have raised severe concerns about the authenticity of online media. Prior works for deepfake detection have made many efforts to capture the intra-modal artifacts. However, deepfake videos in real-world scenarios often consist of a combination of audio and visual. In this paper, we propose an Audio-Visual Joint Learning for Detecting Deepfake (AVoiD-DF), which exploits audio-visual inconsistency for multi-modal forgery detection. Specifically, AVoiD-DF begins by embedding temporal-spatial information in Temporal-Spatial Encoder. A Multi-Modal Joint-Decoder is then designed to fuse multi-modal features and jointly learn inherent relationships. Afterward, a Cross-Modal Classifier is devised to detect manipulation with inter-modal and intra-modal disharmony. Since existing datasets for deepfake detection mainly focus on one modality and only cover a few forgery methods, we build a novel benchmark DefakeAVMiT for multi-modal deepfake detection. DefakeAVMiT contains sufficient visuals with corresponding audios, where any one of the modalities may be maliciously modified by multiple deepfake methods. The experimental results on DefakeAVMiT, FakeAVCeleb, and DFDC demonstrate that the AVoiD-DF outperforms many state-of-the-arts in deepfake detection. Our proposed method also yields superior generalization on various forgery techniques.
Keyword:
Deepfakes
Visualization
Forgery
Detectors
Feature extraction
Faces
Electronic mail
Deepfake detection
multi-modal
audio-visual
joint learning

期刊

IEEE Transactions on Information Forensics and Security 封面图
IEEE Transactions on Information Forensics and Security
IF:
8
论文数:
5.3K
被引数:
2.3W

机构

S
Sun Yat Sen University
学者数:
9.9W
论文数: 7.2W
被引数: 95
P
peking university
学者数:
11.9W
论文数: 8.7W
被引数: 146
J
jinan university
学者数:
4.3W
论文数: 2.7W
被引数: 38
Z
zhejiang university
学者数:
17.7W
论文数: 12.1W
被引数: 152
学者 查看更多机构
引用论文

引用论文

err分享
err收藏
New Emerging Potentials for Human Wharton’s Jelly Mesenchymal Stem Cells: Immunological Features and Hepatocyte-Like Differentiative Capacity
err2010-04-01
err0
PREAI
errRita Anzalone; Melania Lo Iacono; Simona Corrao; Francesca Magno; Tiziana Loria; Francesco Cappello; Giovanni Zummo; Felicia Farina; Giampiero La Rocca
err分享
err收藏
Digital urban production: how does Industry 4.0 reconfigure productive value creation in urban contexts?
err2021-08-19
err0
errOAAI
errHans-Christian Busch; Caroline Mühl; Martina Fuchs; Martina Fromhold-Eisebith
err分享
err收藏
Voice-Face Homogeneity Tells Deepfake
err2023-11-10
err14
errOAAI
errCheng, Harry; Guo, Yangyang; Wang, Tianyi; Li, Qi; Chang, Xiaojun; Nie, Liqiang
err分享
err收藏
err分享
err收藏
学者 查看更多内容