Deep learning for deepfakes creation and detection: A survey

Nguyen, Thanh Thi; Nguyen, Quoc Viet Hung; Nguyen, Dung Tien; Nguyen, Duc Thanh; Huynh‐The, Thien; Nahavandi, Saeid; Nguyên, Thành Tâm; Pham, Quoc-Viet; Nguyen, Cuong M.

doi:10.1016/j.cviu.2022.103525

Cited by 204 publications

(83 citation statements)

References 80 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…However, fake news is widely spread in image and video formats [69]. Deep fake algorithms can create almost real fake images and videos that cannot be comprehended by a human brain as authentic, hence can take the internet by storm by putting deep fakes to serve a propaganda [70]. This is mainly due to the evolution of free and easy-to-use tools for manipulating images and videos which are available in numerous quantities [71].…”

Section: Manipulation Of Deep Fake Algorithms In Images and Videosmentioning

confidence: 99%

Detecting and Mitigating the Dissemination of Fake News: Challenges and Future Research Opportunities

Shahid

Jamshidi

Hakak

et al. 2024

IEEE Trans. Comput. Soc. Syst.

View full text Add to dashboard Cite

Fake news is a major threat to democracy (e.g., influencing public opinion), and its impact cannot be understated particularly in our current socially and digitally connected society. The research community from different disciplines (e.g., computer science, political science, information science, and linguistics) have also studied the dissemination, detection and mitigation of fake news, however it remains challenging to detect and prevent the dissemination of fake news in practice. With AI powered systems, its highly crucial to understand the detector's decision of fake news by means of proper user-friendly explanations when it comes to social media. Hence, in this paper, we systematically survey existing state-of-the-art approaches designed to detect and mitigate the dissemination of fake news, and based on the analysis, we discuss several key challenges and present potential future research agenda specially incorporating AI explainable Fake news credibility system.

show abstract

Section: Manipulation Of Deep Fake Algorithms In Images and Videosmentioning

confidence: 99%

Detecting and Mitigating the Dissemination of Fake News: Challenges and Future Research Opportunities

Shahid

Jamshidi

Hakak

et al. 2024

IEEE Trans. Comput. Soc. Syst.

View full text Add to dashboard Cite

show abstract

“…To mitigate the risk posed by deep fakes, the vision community has developed a series of effective deep fake detection methods [26,11,31] trained on large-scale deepfake datasets. The popular deep fake detection methods include convolutional neural networks (CNN) for detecting visual artifacts [22] and blending boundaries [19], mouth movement analysis [14] and behavioral biometrics [2].…”

Section: Celebdf Distributionmentioning

confidence: 99%

“…Studies have also been proposed for improving the performance of deepfake detectors across datasets and deep fake generation methods using techniques such as reinforcement learning [23] and fine-grained multi-attention network [36]. Readers are referred to the published survey in [31], [26] for detailed information on deep fake detection methods.…”

Section: Deepfake Detectionmentioning

confidence: 99%

“…However, synthetic media generation techniques used by malicious users to deceive pose a severe societal and political threat. In this context, Deepfakes -facial forgery technique that depicts human subjects with altered identities or malicious actions using various deep fake generation techniques-has been flagged as a top AI threat [26,11,31,19,33]. Deep fakes have been used to commit fraud, falsify evidence, manipulate public debates, and destabilize political processes [9,31].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

GBDF: Gender Balanced DeepFake Dataset Towards Fair DeepFake Detection

Nadimpalli¹,

Rattani²

2022

Preprint

View full text Add to dashboard Cite

Facial forgery by deepfakes has raised severe societal concerns. Several solutions have been proposed by the vision community to effectively combat the misinformation on the internet via automated deepfake detection systems. Recent studies have demonstrated that facial analysis-based deep learning models can discriminate based on protected attributes. For the commercial adoption and massive roll-out of the deepfake detection technology, it is vital to evaluate and understand the fairness (the absence of any prejudice or favoritism) of deepfake detectors across demographic variations such as gender and race. As the performance differential of deepfake detectors between demographic subgroups would impact millions of people of the deprived sub-group. This paper aims to evaluate the fairness of the deepfake detectors across males and females. However, existing deepfake datasets are not annotated with demographic labels to facilitate fairness analysis. To this aim, we manually annotated existing popular deepfake datasets with gender labels and evaluated the performance differential of current deepfake detectors across gender. Our analysis on the gender-labeled version of the datasets suggests (a) current deepfake datasets have skewed distribution across gender, and (b) commonly adopted deepfake detectors obtain unequal performance across gender with mostly males outperforming females. Finally, we contributed a gender-balanced and annotated deepfake dataset, GBDF, to mitigate the performance differential and to promote research and development towards fairness-aware deep fake detectors. The GBDF dataset is publicly available at: https://github.com/aakash4305/GBDF

show abstract

“…Mediante un deepfake resulta prácticamente imposible saber si la imagen o la voz de una persona es falsa o verdadera, o si la autenticidad de una fotografía, un sonido o un documento audiovisual se encuentra bajo sospecha de haber sido manipulada o creada artificialmente (Kietzmann et al, 2020). Un deepfake es un producto que combina, superpone, fusiona o reemplaza varios tipos de contenido para producir medios sintéticos con IA y aprendizaje profundo, lo cual ensombrece la noción de autenticidad (Maras y Alexandrou, 2018;Nguyen, 2019).…”

Section: Introductionunclassified

Evolución del Deepfake: campos semánticos y géneros discursivos 2017-2021

Capistrán

2022

icono14

View full text Add to dashboard Cite

El presente estudio tiene como finalidad realizar un análisis de la evolución de los campos semánticos y géneros discursivos del deepfake desde su aparición en 2017 y hasta 2021. El estudio se centra en la descripción de la “capa exterior” (Simondon, 2013) del deepfake como objeto técnico-estético y cultural, en un escenario dinámico, progresivo y cambiante, en donde se expanden sus usos y prácticas culturales. Igualmente se revisan los usos benéficos y delictivos del deepfake, así como los principales problemas éticos y legales que plantea en el complejo escenario de la posverdad. La investigación tiene como base una metodología longitudinal analítica-sintética en tres fases entre 2017-2021. El estudio revela una progresiva evolución discursiva en donde emergen nuevos campos semánticos y géneros discursivos del deepfake, con una tendencia hacia usos benéficos y no sólo delictivos en la esfera de la industria audiovisual, el activismo, el arte experimental, los usos comerciales, médicos, la publicidad, la propaganda, la educación, entre otros; asimismo, el estudio revisa el estado sobre el uso del término deepfake para singificar usos benéficos y delictivos.

show abstract

Deep learning for deepfakes creation and detection: A survey

Cited by 204 publications

References 80 publications

Detecting and Mitigating the Dissemination of Fake News: Challenges and Future Research Opportunities

Detecting and Mitigating the Dissemination of Fake News: Challenges and Future Research Opportunities

GBDF: Gender Balanced DeepFake Dataset Towards Fair DeepFake Detection

Evolución del Deepfake: campos semánticos y géneros discursivos 2017-2021

Contact Info

Product

Resources

About