This work uses adversarial perturbations to enhance deepfake images and fool common deepfake detectors. We created adversarial perturbations using the Fast Gradient Sign Method and the Carlini and Wagner L2 norm attack in both blackbox and whitebox settings. Detectors achieved over 95% accuracy on unperturbed deepfakes, but less than 27% accuracy on perturbed deepfakes. We also explore two improvements to deepfake detectors: (i) Lipschitz regularization, and (ii) Deep Image Prior (DIP). Lipschitz regularization constrains the gradient of the detector with respect to the input in order to increase robustness to input perturbations. The DIP defense removes perturbations using generative convolutional neural networks in an unsupervised manner. Regularization improved the detection of perturbed deepfakes on average, including a 10% accuracy boost in the blackbox case. The DIP defense achieved 95% accuracy on perturbed deepfakes that fooled the original detector, while retaining 98% accuracy in other cases on a 100 image subsample.
Deepfakes utilise Artificial Intelligence (AI) techniques to create synthetic media where the likeness of one person is replaced with another. There are growing concerns that deepfakes can be maliciously used to create misleading and harmful digital contents. As deepfakes become more common, there is a dire need for deepfake detection technology to help spot deepfake media. Present deepfake detection models are able to achieve outstanding accuracy (>90%). However, most of them are limited to within-dataset scenario, where the same dataset is used for training and testing. Most models do not ge…
Better generative models and larger datasets have led to more realistic fake videos that can fool the human eye but produce temporal and spatial artifacts that deep learning approaches can detect. Most current Deepfake detection methods only use individual video frames and therefore fail to learn from temporal information. We created a benchmark of the performance of spatiotemporal convolutional methods using the Celeb-DF dataset. Our methods outperformed state-of-the-art frame-based detection methods. Code for our paper is publicly available at https://github.com/oidelima/Deepfake-Detection.
Adversarial training is widely believed to be a reliable approach to improve model robustness against adversarial attack. However, in this paper, we show that when trained on one type of poisoned data, adversarial training can also be fooled to have catastrophic behavior, e.g., $<1\%$ robust test accuracy with $>90\%$ robust training accuracy on CIFAR-10 dataset. Previously, there are other types of noise poisoned in the training data that have successfully fooled standard training ($15.8\%$ standard test accuracy with $99.9\%$ standard training accuracy on CIFAR-10 dataset), but their poisoni…
Recent advances in video manipulation techniques have made the generation of fake videos more accessible than ever before. Manipulated videos can fuel disinformation and reduce trust in media. Therefore detection of fake videos has garnered immense interest in academia and industry. Recently developed Deepfake detection methods rely on deep neural networks (DNNs) to distinguish AI-generated fake videos from real videos. In this work, we demonstrate that it is possible to bypass such detectors by adversarially modifying fake videos synthesized using existing Deepfake generation methods. We furt…
Adversarially Robust Deepfake Detection via Adversarial Feature Similarity Learning
Research Center for Information Technology Innovation, Academia Sinica（中央研究院資訊科技創新研究中心）；National Chengchi University（國立政治大學）；National Taiwan University（國立臺灣大學）
University of Tübingen（图宾根大学）；Bernstein Center for Computational Neuroscience, Tübingen（图宾根伯恩斯坦计算神经科学中心）；International Max Planck Research School for Intelligent Systems, Tübingen（图宾根国际马克斯·普朗克智能系统研究所）；Max Planck Institute for Biological Cybernetics（马克斯·普朗克生物控制研究所）；Institute for Theoretical Physics, University of Tübingen（图宾根大学理论物理研究所）
Even todays most advanced machine learning models are easily fooled by almost imperceptible perturbations of their inputs. Foolbox is a new Python package to generate such adversarial perturbations and to quantify and compare the robustness of machine learning models. It is build around the idea that the most comparable robustness measure is the minimum perturbation needed to craft an adversarial example. To this end, Foolbox provides reference implementations of most published adversarial attack methods alongside some new ones, all of which perform internal hyperparameter tuning to find the m…
Deepfakes pose growing challenges to the trust of information on the Internet. Thus, detecting deepfakes has attracted increasing attentions from both academia and industry. State-of-the-art deepfake detection methods consist of two key components, i.e., face extractor and face classifier, which extract the face region in an image and classify it to be real/fake, respectively. Existing studies mainly focused on improving the detection performance in non-adversarial settings, leaving security of deepfake detection in adversarial settings largely unexplored. In this work, we aim to bridge the ga…
ByteDance AI Lab（字节跳动人工智能实验室）；Harbin Institute of Technology（哈尔滨工业大学）
With the progress in AI-based facial forgery (i.e., deepfake), people are increasingly concerned about its abuse. Albeit effort has been made for training classification (also known as deepfake detection) models to recognize such forgeries, existing models suffer from poor generalization to unseen forgery technologies and high sensitivity to changes in image/video quality. In this paper, we advocate adversarial training for improving the generalization ability to both unseen facial forgeries and unseen image/video qualities. We believe training with samples that are adversarially crafted to at…
University of Eastern Finland（东芬兰大学）；Institute for Infocomm Research, A*STAR（新加坡科技研究局资讯通信研究院）
Since the invention of cinema, the manipulated videos have existed. But generating manipulated videos that can fool the viewer has been a time-consuming endeavor. With the dramatic improvements in the deep generative modeling, generating believable looking fake videos has become a reality. In the present work, we concentrate on the so-called deepfake videos, where the source face is swapped with the targets. We argue that deepfake detection task should be viewed as a screening task, where the user, such as the video streaming platform, will screen a large number of videos daily. It is clear th…
Facially manipulated images and videos or DeepFakes can be used maliciously to fuel misinformation or defame individuals. Therefore, detecting DeepFakes is crucial to increase the credibility of social media platforms and other media sharing web sites. State-of-the art DeepFake detection techniques rely on neural network based classification models which are known to be vulnerable to adversarial examples. In this work, we study the vulnerabilities of state-of-the-art DeepFake detection methods from a practical stand point. We perform adversarial attacks on DeepFake detectors in a black box set…
Deep neural networks have been shown to exhibit an intriguing vulnerability to adversarial input images corrupted with imperceptible perturbations. However, the majority of adversarial attacks assume global, fine-grained control over the image pixel space. In this paper, we consider a different setting: what happens if the adversary could only alter specific attributes of the input image? These would generate inputs that might be perceptibly different, but still natural-looking and enough to fool a classifier. We propose a novel approach to generate such `semantic' adversarial examples by opti…
Recently it's been shown that neural networks can use images of human faces to accurately predict Body Mass Index (BMI), a widely used health indicator. In this paper we demonstrate that a neural network performing BMI inference is indeed vulnerable to test-time adversarial attacks. This extends test-time adversarial attacks from classification tasks to regression. The application we highlight is BMI inference in the insurance industry, where such adversarial attacks imply a danger of insurance fraud.
Recent advances in autoencoders and generative models have given rise to effective video forgery methods, used for generating so-called "deepfakes". Mitigation research is mostly focused on post-factum deepfake detection and not on prevention. We complement these efforts by introducing a novel class of adversarial attacks---training-resistant attacks---which can disrupt face-swapping autoencoders whether or not its adversarial images have been included in the training set of said autoencoders. We propose the Oscillating GAN (OGAN) attack, a novel attack optimized to be training-resistant, whic…
DeepFake face swapping presents a significant threat to online security and social media, which can replace the source face in an arbitrary photo/video with the target face of an entirely different person. In order to prevent this fraud, some researchers have begun to study the adversarial methods against DeepFake or face manipulation. However, existing works focus on the white-box setting or the black-box setting driven by abundant queries, which severely limits the practical application of these methods. To tackle this problem, we introduce a practical adversarial attack that does not requir…
CYENS Centre of Excellence（CYENS卓越中心）；MBZUAI（穆罕默德·本·扎耶德人工智能大学）
Deepfakes are synthetically generated images, videos or audios, which fraudsters use to manipulate legitimate information. Current deepfake detection systems struggle against unseen data. To address this, we employ three different deep Convolutional Neural Network (CNN) models, (1) VGG16, (2) InceptionV3, and (3) XceptionNet to classify fake and real images extracted from videos. We also constructed a fusion of the deep CNN models to improve the robustness and generalisation capability. The proposed technique outperforms state-of-the-art models with 96.5% accuracy, when tested on publicly avai…
Face modification systems using deep learning have become increasingly powerful and accessible. Given images of a person's face, such systems can generate new images of that same person under different expressions and poses. Some systems can also modify targeted attributes such as hair color or age. This type of manipulated images and video have been coined Deepfakes. In order to prevent a malicious user from generating modified images of a person without their consent we tackle the new problem of generating adversarial attacks against such image translation systems, which disrupt the resultin…
Kyushu University（九州大学）；Advanced Telecommunications Research Institute International（先进电信研究国际研究所）
Recent research has revealed that the output of Deep Neural Networks (DNN) can be easily altered by adding relatively small perturbations to the input vector. In this paper, we analyze an attack in an extremely limited scenario where only one pixel can be modified. For that we propose a novel method for generating one-pixel adversarial perturbations based on differential evolution (DE). It requires less adversarial information (a black-box attack) and can fool more types of networks due to the inherent features of DE. The results show that 67.97% of the natural images in Kaggle CIFAR-10 test d…
Recent advancements in CNNs have shown remarkable achievements in various CV/AI applications. Though CNNs show near human or better than human performance in many critical tasks, they are quite vulnerable to adversarial attacks. These attacks are potentially dangerous in real-life deployments. Though there have been many adversarial attacks proposed in recent years, there is no proper way of quantifying the effectiveness of these attacks. As of today, mere fooling rate is used for measuring the susceptibility of the models, or the effectiveness of adversarial attacks. Fooling rate just conside…
University of Illinois at Urbana Champaign（伊利诺伊大学厄巴纳-香槟分校）
An adversarial example is an example that has been adjusted to produce the wrong label when presented to a system at test time. If adversarial examples existed that could fool a detector, they could be used to (for example) wreak havoc on roads populated with smart vehicles. Recently, we described our difficulties creating physical adversarial stop signs that fool a detector. More recently, Evtimov et al. produced a physical adversarial stop sign that fools a proxy model of a detector. In this paper, we show that these physical adversarial stop signs do not fool two standard detectors (YOLO an…
Almost all adversarial attacks are formulated to add an imperceptible perturbation to an image in order to fool a model. Here, we consider the opposite which is adversarial examples that can fool a human but not a model. A large enough and perceptible perturbation is added to an image such that a model maintains its original decision, whereas a human will most likely make a mistake if forced to decide (or opt not to decide at all). Existing targeted attacks can be reformulated to synthesize such adversarial examples. Our proposed attack, dubbed NKE, is similar in essence to the fooling images,…
The University of Melbourne（墨尔本大学）；Peking University（北京大学）；Deakin University（迪肯大学）
Understanding the actions of both humans and artificial intelligence (AI) agents is important before modern AI systems can be fully integrated into our daily life. In this paper, we show that, despite their current huge success, deep learning based AI systems can be easily fooled by subtle adversarial noise to misinterpret the intention of an action in interaction scenarios. Based on a case study of skeleton-based human interactions, we propose a novel adversarial attack on interactions, and demonstrate how DNN-based interaction models can be tricked to predict the participants' reactions in u…
At present, adversarial attacks are designed in a task-specific fashion. However, for downstream computer vision tasks such as image captioning, image segmentation etc., the current deep learning systems use an image classifier like VGG16, ResNet50, Inception-v3 etc. as a feature extractor. Keeping this in mind, we propose Mimic and Fool, a task agnostic adversarial attack. Given a feature extractor, the proposed attack finds an adversarial image which can mimic the image feature of the original image. This ensures that the two images give the same (or similar) output regardless of the task. W…
Indian Institute of Technology, Hyderabad（印度理工学院海得拉巴分校）
Current SOTA adversarially robust models are mostly based on adversarial training (AT) and differ only by some regularizers either at inner maximization or outer minimization steps. Being repetitive in nature during the inner maximization step, they take a huge time to train. We propose a non-iterative method that enforces the following ideas during training. Attribution maps are more aligned to the actual object in the image for adversarially robust models compared to naturally trained models. Also, the allowed set of pixels to perturb an image (that changes model decision) should be restrict…
Neural networks are vulnerable to adversarial examples, malicious inputs crafted to fool trained models. Adversarial examples often exhibit black-box transfer, meaning that adversarial examples for one model can fool another model. However, adversarial examples may be overfit to exploit the particular architecture and feature representation of a source model, resulting in sub-optimal black-box transfer attacks to other target models. This leads us to introduce the Intermediate Level Attack (ILA), which attempts to fine-tune an existing adversarial example for greater black-box transferability …
Fool the Stoplight: Realistic Adversarial Patch Attacks on Traffic Light Detectors
Karlsruhe Institute of Technology (KIT)（卡尔斯鲁厄理工学院）；Department of Technical Cognitive Systems（技术认知系统系）；FZI Research Center for Information Technology（信息技术研究所以）
The University of Adelaide（阿德莱德大学）；Tencent AI Lab（腾讯人工智能实验室）
Recent studies in deepfake detection have yielded promising results when the training and testing face forgeries are from the same dataset. However, the problem remains challenging when one tries to generalize the detector to forgeries created by unseen methods in the training dataset. This work addresses the generalizable deepfake detection from a simple principle: a generalizable representation should be sensitive to diverse types of forgeries. Following this principle, we propose to enrich the "diversity" of forgeries by synthesizing augmented forgeries with a pool of forgery configurations…
Paulson School of Engineering and Applied Sciences, Harvard University（哈佛大学保尔森工程与应用科学学院）
Deep neural networks have proven remarkably effective at solving many classification problems, but have been criticized recently for two major weaknesses: the reasons behind their predictions are uninterpretable, and the predictions themselves can often be fooled by small adversarial perturbations. These problems pose major obstacles for the adoption of neural networks in domains that require security or transparency. In this work, we evaluate the effectiveness of defenses that differentiably penalize the degree to which small changes in inputs can alter model predictions. Across multiple atta…
Deepfakes raised serious concerns on the authenticity of visual contents. Prior works revealed the possibility to disrupt deepfakes by adding adversarial perturbations to the source data, but we argue that the threat has not been eliminated yet. This paper presents MagDR, a mask-guided detection and reconstruction pipeline for defending deepfakes from adversarial attacks. MagDR starts with a detection module that defines a few criteria to judge the abnormality of the output of deepfakes, and then uses it to guide a learnable reconstruction procedure. Adaptive masks are extracted to capture the…
Evading DeepFake Detectors via Adversarial Statistical Consistency
Kyushu University（九州大学）；Centre for Frontier AI Research (CFAR), A*STAR（新加坡科技研究局前沿人工智能研究中心）；Institute of High Performance Computing (IHPC), A*STAR（新加坡科技研究局高性能计算研究所）；Nanyang Technological University（南洋理工大学）；Singapore Management University（新加坡管理大学）；University of Alberta（阿尔伯塔大学）；The University of Tokyo（东京大学）
Existing pixel-level adversarial attacks on neural networks may be deficient in real scenarios, since pixel-level changes on the data cannot be fully delivered to the neural network after camera capture and multiple image preprocessing steps. In contrast, in this paper, we argue from another perspective that gaussian blurring, a common technique of image preprocessing, can be aggressive itself in specific occasions, thus exposing the network to real-world adversarial attacks. We first propose an adversarial attack demo named peak suppression (PS) by suppressing the values of peak elements in t…
DeepForgeSeal: Latent Space-Driven Semi-Fragile Watermarking for Deepfake Detection Using Adversarial Reinforcement Learning
The Signal Processing, Artificial Intelligence and Vision Technologies (SAIVT), Queensland University of Technology, Australia（信号处理、人工智能与视觉技术研究所（SAIVT），昆士兰理工大学）
Deepfake videos, where a person's face is automatically swapped with a face of someone else, are becoming easier to generate with more realistic results. In response to the threat such manipulations can pose to our trust in video evidence, several large datasets of deepfake videos and many methods to detect them were proposed recently. However, it is still unclear how realistic deepfake videos are for an average person and whether the algorithms are significantly better than humans at detecting them. In this paper, we present a subjective study conducted in a crowdsourcing-like scenario, which…
Adversarial attacks on machine learning models have seen increasing interest in the past years. By making only subtle changes to the input of a convolutional neural network, the output of the network can be swayed to output a completely different result. The first attacks did this by changing pixel values of an input image slightly to fool a classifier to output the wrong class. Other approaches have tried to learn "patches" that can be applied to an object to fool detectors and classifiers. Some of these approaches have also shown that these attacks are feasible in the real-world, i.e. by mod…
Despite high accuracy of Convolutional Neural Networks (CNNs), they are vulnerable to adversarial and out-distribution examples. There are many proposed methods that tend to detect or make CNNs robust against these fooling examples. However, most such methods need access to a wide range of fooling examples to retrain the network or to tune detection parameters. Here, we propose a method to detect adversarial and out-distribution examples against a pre-trained CNN without needing to retrain the CNN or needing access to a wide variety of fooling examples. To this end, we create adversarial profi…
Recent works have demonstrated the existence of {\it adversarial examples} targeting a single machine learning system. In this paper we ask a simple but fundamental question of "selective fooling": given {\it multiple} machine learning systems assigned to solve the same classification problem and taking the same input signal, is it possible to construct a perturbation to the input signal that manipulates the outputs of these {\it multiple} machine learning systems {\it simultaneously} in arbitrary pre-defined ways? For example, is it possible to selectively fool a set of "enemy" machine learni…
The Partnership on AI（人工智能合作组织）；XPRIZE Foundation（XPRIZE基金会）；Syntiant Corp（Syntiant公司）；Thoughtful Technology Project（深思技术项目）
Synthetic media detection technologies label media as either synthetic or non-synthetic and are increasingly used by journalists, web platforms, and the general public to identify misinformation and other forms of problematic content. As both well-resourced organizations and the non-technical general public generate more sophisticated synthetic media, the capacity for purveyors of problematic content to adapt induces a \newterm{detection dilemma}: as detection practices become more accessible, they become more easily circumvented. This paper describes how a multistakeholder cohort from academi…
A variety of effective face-swap and face-reenactment methods have been publicized in recent years, democratizing the face synthesis technology to a great extent. Videos generated as such have come to be called deepfakes with a negative connotation, for various social problems they have caused. Facing the emerging threat of deepfakes, we have built the Korean DeepFake Detection Dataset (KoDF), a large-scale collection of synthesized and real videos focused on Korean subjects. In this paper, we provide a detailed description of methods used to construct the dataset, experimentally show the disc…
Pohang University of Science and Technology (POSTECH)（浦项科技大学）
Adversarial examples are inputs intentionally generated for fooling a deep neural network. Recent studies have proposed unrestricted adversarial attacks that are not norm-constrained. However, the previous unrestricted attack methods still have limitations to fool real-world applications in a black-box setting. In this paper, we present a novel method for generating unrestricted adversarial examples using GAN where an attacker can only access the top-1 final decision of a classification model. Our method, Latent-HSJA, efficiently leverages the advantages of a decision-based attack in the laten…
