DeepFake video production and SIFT-based analysis

Đorđević, Miljan; Milivojević, Milan S.; Gavrovska, Ana

doi:10.5937/telfor2001022q

Cited by 2 publications

(2 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…e face detector in the Dlib library is used to extract the images of the face region, which are used as the inputs for model training, validation, and testing. For the models, we use the Xception [4] as the target model for attack and defense and the MesoInception [5] as the substitute model for the Xception.…”

Section: Experiments Settingmentioning

confidence: 99%

See 1 more Smart Citation

A Novel Defensive Strategy for Facial Manipulation Detection Combining Bilateral Filtering and Joint Adversarial Training

Luo

Feng

Weng

et al. 2021

Security and Communication Networks

View full text Add to dashboard Cite

Facial manipulation enables facial expressions to be tampered with or facial identities to be replaced in videos. The fake videos are so realistic that they are even difficult for human eyes to distinguish. This poses a great threat to social and public information security. A number of facial manipulation detectors have been proposed to address this threat. However, previous studies have shown that the accuracy of these detectors is sensitive to adversarial examples. The existing defense methods are very limited in terms of applicable scenes and defense effects. This paper proposes a new defense strategy for facial manipulation detectors, which combines a passive defense method, bilateral filtering, and a proactive defense method, joint adversarial training, to mitigate the vulnerability of facial manipulation detectors against adversarial examples. The bilateral filtering method is applied in the preprocessing stage of the model without any modification to denoise the input adversarial examples. The joint adversarial training starts from the training stage of the model, which mixes various adversarial examples and original examples to train the model. The introduction of joint adversarial training can train a model that defends against multiple adversarial attacks. The experimental results show that the proposed defense strategy positively helps facial manipulation detectors counter adversarial examples.

show abstract

Section: Experiments Settingmentioning

confidence: 99%

“…Facial manipulation detectors can be grouped into two broad categories. One is based on the manual feature extraction [3][4][5], and the other is on various deep neural networks [6][7][8][9][10]. Compared with traditional feature extraction methods, deep neural network-based methods generally have better detection performance.…”

Section: Introductionmentioning

confidence: 99%

A Novel Defensive Strategy for Facial Manipulation Detection Combining Bilateral Filtering and Joint Adversarial Training

Luo

Feng

Weng

et al. 2021

Security and Communication Networks

View full text Add to dashboard Cite

show abstract

An Enhanced Deep Learning-Based DeepFake Video Detection and Classification System

et al. 2022

View full text Add to dashboard Cite

The privacy of individuals and entire countries is currently threatened by the widespread use of face-swapping DeepFake models, which result in a sizable number of fake videos that seem extraordinarily genuine. Because DeepFake production tools have advanced so much and since so many researchers and businesses are interested in testing their limits, fake media is spreading like wildfire over the internet. Therefore, this study proposes five-layered convolutional neural networks (CNNs) for a DeepFake detection and classification model. The CNN enhanced with ReLU is used to extract features from these faces once the model has extracted the face region from video frames. To guarantee model accuracy while maintaining a suitable weight, a CNN enabled with ReLU model was used for the DeepFake-detection-influenced video. The performance evaluation of the proposed model was tested using Face2Face, and first-order motion DeepFake datasets. Experimental results revealed that the proposed model has an average prediction rate of 98% for DeepFake videos and 95% for Face2Face videos under actual network diffusion circumstances. When compared with systems such as Meso4, MesoInception4, Xception, EfficientNet-B0, and VGG16 which utilizes the convolutional neural network, the suggested model produced the best results with an accuracy rate of 86%.

show abstract

DeepFake video production and SIFT-based analysis

Cited by 2 publications

References 9 publications

A Novel Defensive Strategy for Facial Manipulation Detection Combining Bilateral Filtering and Joint Adversarial Training

A Novel Defensive Strategy for Facial Manipulation Detection Combining Bilateral Filtering and Joint Adversarial Training

An Enhanced Deep Learning-Based DeepFake Video Detection and Classification System

Contact Info

Product

Resources

About