Data Augmentation-based Novel Deep Learning Method for Deepfaked Images Detection

被引:12
作者
Iqbal, Farkhund [1 ]
Abbasi, Ahmed [2 ]
Javed, Abdul rehman [3 ]
Almadhor, Ahmad [4 ]
Jalil, Zunera [5 ]
Anwar, Sajid [6 ]
Rida, Imad [7 ]
机构
[1] Zayed Univ, Coll Technol Innovat, Abu Dhabi 144534, U Arab Emirates
[2] Air Univ, Dept Creat Technol, PAF Complex, Islamabad 56300, Pakistan
[3] Lebanese Amer Univ, Dept Elect & Comp Engn, Byblos, Lebanon
[4] Jouf Univ, Coll Comp & Informat Sci, Dept Comp Engn & Networks, Sakaka 42421, Saudi Arabia
[5] Air Univ, Dept Cyber Secur, PAF Complex, Islamabad 56300, Pakistan
[6] Inst Management Sci Peshawar, Peshawar 25130, Pakistan
[7] Univ Technol Compiegne, BMBI Lab, F-60203 Compiegne, France
关键词
Deepfake detection; data augmentation; image processing; deep learning; artificial intelligence; transfer learning;
D O I
10.1145/3592615
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Recent advances in artificial intelligence have led to deepfake images, enabling users to replace a real face with a genuine one. deepfake images have recently been used to malign public figures, politicians, and even average citizens. deepfake but realistic images have been used to stir political dissatisfaction, blackmail, propagate false news, and even carry out bogus terrorist attacks. Thus, identifying real images from fakes has got more challenging. To avoid these issues, this study employs transfer learning and data augmentation technique to classify deepfake images. For experimentation, 190,335 RGB-resolution deepfake and real images and image augmentation methods are used to prepare the dataset. The experiments use the deep learning models: convolutional neural network (CNN), Inception V3, visual geometry group (VGG19), and VGG16 with a transfer learning approach. Essential evaluation metrics (accuracy, precision, recall, F1-score, confusion matrix, and AUC-ROC curve score) are used to test the efficacy of the proposed approach. Results revealed that the proposed approach achieves an accuracy, recall, F1-score and AUC-ROC score of 90% and 91% precision, with our fine-tuned VGG16 model outperforming other DL models in recognizing real and deepfakes.
引用
收藏
页数:15
相关论文
共 58 条
[1]  
Passos LA, 2022, Arxiv, DOI arXiv:2202.06095
[2]   Deep learning for religious and continent-based toxic content detection and classification [J].
Abbasi, Ahmed ;
Javed, Abdul Rehman ;
Iqbal, Farkhund ;
Kryvinska, Natalia ;
Jalil, Zunera .
SCIENTIFIC REPORTS, 2022, 12 (01)
[3]   A Large-Scale Benchmark Dataset for Anomaly Detection and Rare Event Classification for Audio Forensics [J].
Abbasi, Ahmed ;
Javed, Abdul Rehman Rehman ;
Yasin, Amanullah ;
Jalil, Zunera ;
Kryvinska, Natalia ;
Tariq, Usman .
IEEE ACCESS, 2022, 10 :38885-38894
[4]  
Afchar D, 2018, IEEE INT WORKS INFOR
[5]  
Ahn YC, 2019, INT JOINT CONF COMP, P289, DOI [10.1109/jcsse.2019.8864171, 10.1109/JCSSE.2019.8864171]
[6]   Low-Light Image Enhancement Using Image-to-Frequency Filter Learning [J].
Al Sobbahi, Rayan ;
Tekli, Joe .
IMAGE ANALYSIS AND PROCESSING, ICIAP 2022, PT II, 2022, 13232 :693-705
[7]   Low-Light Homomorphic Filtering Network for integrating image enhancement and classification [J].
Al Sobbahi, Rayan ;
Tekli, Joe .
SIGNAL PROCESSING-IMAGE COMMUNICATION, 2022, 100
[8]   Examining an integrated model of green image, perceived quality, satisfaction, trust, and loyalty in upscale hotels [J].
Assaker, Guy ;
O'Connor, Peter ;
El-Haddad, Rania .
JOURNAL OF HOSPITALITY MARKETING & MANAGEMENT, 2020, 29 (08) :934-955
[9]  
Chang X, 2020, CHIN CONTR CONF, P7252, DOI [10.23919/CCC50068.2020.9189596, 10.23919/ccc50068.2020.9189596]
[10]   FSSPOTTER: SPOTTING FACE-SWAPPED VIDEO BY SPATIAL AND TEMPORAL CLUES [J].
Chen, Peng ;
Liu, Jin ;
Liang, Tao ;
Zhou, Guangzhi ;
Gao, Hongchao ;
Dai, Jiao ;
Han, Jizhong .
2020 IEEE INTERNATIONAL CONFERENCE ON MULTIMEDIA AND EXPO (ICME), 2020,