Deep learning for deepfakes creation and detection: A survey

By Thanh Thi Nguyen, Quoc Viet Hung Nguyen, Dung Nguyen, Dinh C. Nguyen, Thien Huynh-The, Saeid Nahavandi, Thanh Tam Nguyen, Quoc-Viet Pham, Cu Nguyen. Computer Vision and Image Understanding, 2019

629 citations recorded by the source corpus.

Deep learning has been successfully applied to solve various complex problems ranging from big data analytics to computer vision and human-level control. Deep learning advances however have also been employed to create software that can cause threats to privacy, democracy and national security. One of those deep learning-powered applications recently emerged is "deepfake". Deepfake algorithms can create fake images and videos that humans cannot distinguish them from authentic ones. The proposal of technologies that can automatically detect and assess the integrity of digital visual media is therefore indispensable. This paper presents a survey of algorithms used to create deepfakes and, more importantly, methods proposed to detect deepfakes in the literature to date. We present extensive discussions on challenges, research trends and directions related to deepfake technologies. By reviewing the background of deepfakes and state-of-the-art deepfake detection methods, this study provides a comprehensive overview of deepfake techniques and facilitates the development of new and more robust methods to deal with the increasingly challenging deepfakes.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/86a0919e924389260a47b492f35f9761fcd3433d

How researchers see this paper

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

In Computer Vision and Pattern Recognition

Citations in the corpus

EditGAN: High-Precision Semantic Image Editing (2021, methodology): “[84] Thanh Thi Nguyen, Quoc Viet Hung Nguyen, Cuong M.”
Towards Measuring Fairness in AI: The Casual Conversations Dataset (2021, background): “Fake content generation has rapidly evolved in the past several years and recent studies [18] show that the gap between the deepfake generators and detectors has been growing quickly.”
Behavioral Use Licensing for Responsible AI (2020, background): “Recent progress in generative models, such as those based on Generative Adversarial Networks (GANs) for images [32], and Transformer Architectures [64] for text, have to led to fears of these models being exploited for malicious use via Deep Fakes [47], or the creation of machine-generated fake news”
Deep Reinforcement Learning for Cyber Security (2019, background): “, deepfakes [183]), hackers can spread false news in elections or manipulate financial markets [184].”
Character-LLM: A Trainable Agent for Role-Playing (2023, background): “Besides languages, multimodal simulacra of humankind are also studied such as voice generation (Wang et al., 2023a; Zhang et al., 2023) or deepfakes creation (Wang et al., 2021; Nguyen et al., 2022).”
LION: Latent Point Diffusion Models for 3D Shape Generation (2022, background): “[123] Thanh Thi Nguyen, Quoc Viet Hung Nguyen, Cuong M.”
A Statistical Turing Test for Generative Models (2023, background): “See surveys [43, 29] for detailed descriptions of state-of-the-art generation and detection methods in vision.”