Protecting World Leaders Against Deep Fakes

By Shruti Agarwal, Hany Farid, Yuming Gu, Mingming He, Koki Nagano, Hao Li. 2019

575 citations recorded by the source corpus.

Subfields: Safety, Robustness & Interpretability

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Safety, Robustness & Interpretability

Citations in the corpus

Diffusion Models Beat GANs on Image Synthesis (2021, background): “images are a more direct concern, and detecting and mitigating propaganda and fake news based on generative models is an ongoing area of research [4, 3, 5].”
FakeCatcher: Detection of Synthetic Portrait Videos using Biological Signals (2019, background): “hreat before having serious consequences. Although there are recent purely deep learning approaches to detect fake content, those are limited by the speciﬁc generative model [4], dataset [56], people [5], or hand-crafted features [42]. In contrast to all, we choose to search for some natural priors ”
Media Forensics and DeepFakes: An Overview (2020, unknown): “A different approach is followed in [170].”
A Continual Deepfake Detection Benchmark: Dataset, Methods, and Essentials (2022, background): “To address this issue, many deepfake detection datasets (e.g.,[40, 44, 46, 21, 64, 19, 31]) and techniques (e.g., [87, 4, 80, 57, 56, 5]) are proposed.”
Preserving integrity in online social networks (2020, background): “Within this same field, motion cues associated with target individuals [10] are very distinctive and can be used as a soft biometric to assess the identity of someone, e.g., how someone smiles or talks.”
Face X-Ray for More General Face Forgery Detection (2019, background): “For instance, a face forensic approach exploiting facial expressions and head movements customized for specific individuals is proposed in [4].”
A review of ensemble learning and data augmentation models for class imbalanced problems: combination, implementation and evaluation (2023, background): “Moreover, GANs [330, 253] and re-lated data generation methods such as deepfake [331, 332, 333] have negative connotations since they can be used to imperson-ate individuals, by generating images, speech and related bio-metrics [334].”
The Creation and Detection of Deepfakes (2020, background): “For example, in [6] the authors protect world leaders from a wide variety of deepfake aacks by modeling their recorded stock footage.”
Two-branch Recurrent Network for Isolating Deepfakes in Videos (2020, methodology): “built person-specific classifiers [7] using one-class support vector machines (SVM) and features computed from Action Units (AU) and 3D head pose movements.”
Detecting and Simulating Artifacts in GAN Fake Images (2019, background): “In response, the forensics community has been working on detecting such generated content [4], [12], [13], [14].”
Deepfake Video Detection Using Convolutional Vision Transformer (2021, background): “It is most popularly used to insert famous actors in a variety of movie clips [2].”
Deepfakes generation and detection: state-of-the-art, open challenges, countermeasures, and way forward (2021, methodology): “Major improvements in detection algorithms include i) identification of artifacts left during the generation process, such as inconsistencies in head pose [71], lack of eye blinking [77], color variations in facial texture [155] and teeth alignment, ii) detection of unseen GAN generated samples, iii”
Exploring Temporal Coherence for More General Video Face Forgery Detection (2021, background): “Some recent studies [8, 4, 46, 14, 27, 56] aim to locate the visual artifacts in the forged images and make predictions based on the location results.”
DeepFakes and Beyond: A Survey of Face Manipulation and Fake Detection (2020, unknown): “Agarwal and Farid proposed in [59] a detection technique based on facial expressions and head movements.”
ContraGAN: Contrastive Learning for Conditional Image Generation (2020, methodology): “We can do so many beneficial tasks with conditional GANs, but we should be concerned that conditional GANs can be used for deepfake techniques [59].”
Deep learning for deepfakes creation and detection: A survey (2019, methodology): “person (master) sitting in front of a camera [1].”
Celeb-DF: A Large-Scale Challenging Dataset for DeepFake Forensics (2019, unknown): “In [7], the idiosyncratic behavioral patterns of a particular individual are captured by the time series of facial landmarks extracted from real videos are used to spot DeepFake videos.”
Creating, Using, Misusing, and Detecting Deep Fakes (2022, background): “…is based on the assumption that as a person speaks, they have distinct facial expressions and movements that can be used to distinguish them from others, impersonators, and deep fakes (Agarwal et al. 2019; Agarwal, Farid, El-Gaaly, et al. 2020; Agarwal et al. 2021; Cozzolino et al. 2021).”