Biblio

In recent years, differential privacy has gradually become a standard definition in the field of data privacy protection. Differential privacy does not need to make assumptions about the prior knowledge of privacy adversaries, so it has a more stringent effect than existing privacy protection models and definitions. This good feature has been used by researchers to solve the in-depth learning problem restricted by the problem of privacy and security, making an important breakthrough, and promoting its further large-scale application. Combining differential privacy with BEGAN, we propose the DP-BEGAN framework. The differential privacy is realized by adding carefully designed noise to the gradient of Gan model training, so as to ensure that Gan can generate unlimited synthetic data that conforms to the statistical characteristics of source data and does not disclose privacy. At the same time, it is compared with the existing methods on public datasets. The results show that under a certain privacy budget, this method can generate higher quality privacy protection data more efficiently, which can be used in a variety of data analysis tasks. The privacy loss is independent of the amount of synthetic data, so it can be applied to large datasets.

CAPTCHA or Completely Automated Public Turing test to Tell Computers and Humans Apart is a technique to distinguish between humans and computers by generating and evaluating tests that can be passed by humans but not computer bots. However, captchas are not foolproof, and they can be bypassed which raises security concerns. Hence, sites over the internet remain open to such vulnerabilities. This research paper identifies the vulnerabilities found in some of the commonly used captcha schemes by cracking them using Deep Learning techniques. It also aims to provide solutions to safeguard against these vulnerabilities and provides recommendations for the generation of secure captchas.

Image forensics is an increasingly relevant problem, as it can potentially address online disinformation campaigns and mitigate problematic aspects of social media. Of particular interest, given its recent successes, is the detection of imagery produced by Generative Adversarial Networks (GANs), e.g. `deepfakes'. Leveraging large training sets and extensive computing resources, recent GANs can be trained to generate synthetic imagery which is (in some ways) indistinguishable from real imagery. We analyze the structure of the generating network of a popular GAN implementation [1], and show that the network's treatment of exposure is markedly different from a real camera. We further show that this cue can be used to distinguish GAN-generated imagery from camera imagery, including effective discrimination between GAN imagery and real camera images used to train the GAN.

Recent advances in technology have made the deep learning (DL) models available for use in a wide variety of novel applications; for example, generative adversarial network (GAN) models are capable of producing hyper-realistic images, speech, and even videos, such as the so-called “Deepfake” produced by GANs with manipulated audio and/or video clips, which are so realistic as to be indistinguishable from the real ones in human perception. Aside from innovative and legitimate applications, there are numerous nefarious or unlawful ways to use such counterfeit contents in propaganda, political campaigns, cybercrimes, extortion, etc. To meet the challenges posed by Deepfake multimedia, we propose a deep ensemble learning technique called DeepfakeStack for detecting such manipulated videos. The proposed technique combines a series of DL based state-of-art classification models and creates an improved composite classifier. Based on our experiments, it is shown that DeepfakeStack outperforms other classifiers by achieving an accuracy of 99.65% and AUROC of 1.0 score in detecting Deepfake. Therefore, our method provides a solid basis for building a Realtime Deepfake detector.

In this paper, we present a semi-supervised remote sensing change detection method based on graph model with Generative Adversarial Networks (GANs). Firstly, the multi-temporal remote sensing change detection problem is converted as a problem of semi-supervised learning on graph where a majority of unlabeled nodes and a few labeled nodes are contained. Then, GANs are adopted to generate samples in a competitive manner and help improve the classification accuracy. Finally, a binary change map is produced by classifying the unlabeled nodes to a certain class with the help of both the labeled nodes and the unlabeled nodes on graph. Experimental results carried on several very high resolution remote sensing image data sets demonstrate the effectiveness of our method.

Despite the popularity of wireless sensor networks (WSNs) in a wide range of applications, security problems associated with them have not been completely resolved. Middleware is generally introduced as an intermediate layer between WSNs and the end user to resolve some limitations, but most of the existing middleware is unable to protect data from malicious and unknown attacks during transmission. This paper introduces an intelligent middleware based on an unsupervised learning technique called Generative Adversarial Networks (GANs) algorithm. GANs contain two networks: a generator (G) network and a detector (D) network. The G creates fake data similar to the real samples and combines it with real data from the sensors to confuse the attacker. The D contains multi-layers that have the ability to differentiate between real and fake data. The output intended for this algorithm shows an actual interpretation of the data that is securely communicated through the WSN. The framework is implemented in Python with experiments performed using Keras. Results illustrate that the suggested algorithm not only improves the accuracy of the data but also enhances its security by protecting data from adversaries. Data transmission from the WSN to the end user then becomes much more secure and accurate compared to conventional techniques.

The problem of network representation learning, also known as network embedding, arises in many machine learning tasks assuming that there exist a small number of variabilities in the vertex representations which can capture the "semantics" of the original network structure. Most existing network embedding models, with shallow or deep architectures, learn vertex representations from the sampled vertex sequences such that the low-dimensional embeddings preserve the locality property and/or global reconstruction capability. The resultant representations, however, are difficult for model generalization due to the intrinsic sparsity of sampled sequences from the input network. As such, an ideal approach to address the problem is to generate vertex representations by learning a probability density function over the sampled sequences. However, in many cases, such a distribution in a low-dimensional manifold may not always have an analytic form. In this study, we propose to learn the network representations with adversarially regularized autoencoders (NetRA). NetRA learns smoothly regularized vertex representations that well capture the network structure through jointly considering both locality-preserving and global reconstruction constraints. The joint inference is encapsulated in a generative adversarial training process to circumvent the requirement of an explicit prior distribution, and thus obtains better generalization performance. We demonstrate empirically how well key properties of the network structure are captured and the effectiveness of NetRA on a variety of tasks, including network reconstruction, link prediction, and multi-label classification.

Hashing has been a widely-adopted technique for nearest neighbor search in large-scale image retrieval tasks. Recent research has shown that leveraging supervised information can lead to high quality hashing. However, the cost of annotating data is often an obstacle when applying supervised hashing to a new domain. Moreover, the results can suffer from the robustness problem as the data at training and test stage may come from different distributions. This paper studies the exploration of generating synthetic data through semi-supervised generative adversarial networks (GANs), which leverages largely unlabeled and limited labeled training data to produce highly compelling data with intrinsic invariance and global coherence, for better understanding statistical structures of natural data. We demonstrate that the above two limitations can be well mitigated by applying the synthetic data for hashing. Specifically, a novel deep semantic hashing with GANs (DSH-GANs) is presented, which mainly consists of four components: a deep convolution neural networks (CNN) for learning image representations, an adversary stream to distinguish synthetic images from real ones, a hash stream for encoding image representations to hash codes and a classification stream. The whole architecture is trained end-to-end by jointly optimizing three losses, i.e., adversarial loss to correct label of synthetic or real for each sample, triplet ranking loss to preserve the relative similarity ordering in the input real-synthetic triplets and classification loss to classify each sample accurately. Extensive experiments conducted on both CIFAR-10 and NUS-WIDE image benchmarks validate the capability of exploiting synthetic images for hashing. Our framework also achieves superior results when compared to state-of-the-art deep hash models.

This work describes how automated data generation integrates in a big data pipeline. A lack of veracity in big data can cause models that are inaccurate, or biased by trends in the training data. This can lead to issues as a pipeline matures that are difficult to overcome. This work describes the use of a Generative Adversarial Network to generate sketch data, such as those that might be used in a human verification task. These generated sketches are verified as recognizable using a crowd-sourcing methodology, and finds that the generated sketches were correctly recognized 43.8% of the time, in contrast to human drawn sketches which were 87.7% accurate. This method is scalable and can be used to generate realistic data in many domains and bootstrap a dataset used for training a model prior to deployment.