Biblio

In this work, we applied a deep Convolutional Neural Network (CNN) with Xception model to perform malware image classification. The Xception model is a recently developed special CNN architecture that is more powerful with less over- fitting problems than the current popular CNN models such as VGG16. However only a few use cases of the Xception model can be found in literature, and it has never been used to solve the malware classification problem. The performance of our approach was compared with other methods including KNN, SVM, VGG16 etc. The experiments on two datasets (Malimg and Microsoft Malware Dataset) demonstrated that the Xception model can achieve the highest training accuracy than all other approaches including the champion approach, and highest validation accuracy than all other approaches including VGG16 model which are using image-based malware classification (except the champion solution as this information was not provided). Additionally, we proposed a novel ensemble model to combine the predictions from .bytes files and .asm files, showing that a lower logloss can be achieved. Although the champion on the Microsoft Malware Dataset achieved a bit lower logloss, our approach does not require any features engineering, making it more effective to adapt to any future evolution in malware, and very much less time consuming than the champion's solution.

With the continuous development of the Internet of Vehicles, vehicles are no longer isolated nodes, but become a node in the car network. The open Internet will introduce traditional security issues into the Internet of Things. In order to ensure the safety of the networked cars, we hope to set up an intrusion detection system (IDS) on the vehicle terminal to detect and intercept network attacks. In our work, we designed an intrusion detection system for the Internet of Vehicles based on a convolutional neural network, which can run in a low-powered embedded vehicle terminal to monitor the data in the car network in real time. Moreover, for the case of packet encryption in some car networks, we have also designed a separate version for intrusion detection by analyzing the packet header. Experiments have shown that our system can guarantee high accuracy detection at low latency for attack traffic.

The use of biometric-based authentication systems spread over daily life consumer electronics. Over the years, researchers' interest shifted from hard (such as fingerprints, voice and keystroke dynamics) to soft biometrics (such as age, ethnicity and gender), mainly by using the latter to improve the authentication systems effectiveness. While newer approaches are constantly being proposed by domain experts, in the last years Deep Learning has raised in many computer vision tasks, also becoming the current state-of-art for several biometric approaches. However, since the automatic processing of data rich in sensitive information could expose users to privacy threats associated to their unfair use (i.e. gender or ethnicity), in the last years researchers started to focus on the development of defensive strategies in the view of a more secure and private AI. The aim of this work is to exploit Adversarial Perturbation, namely approaches able to mislead state-of-the-art CNNs by injecting a suitable small perturbation over the input image, to protect subjects against unwanted soft biometrics-based identification by automatic means. In particular, since ethnicity is one of the most critical soft biometrics, as a case of study we will focus on the generation of adversarial stickers that, once printed, can hide subjects ethnicity in a real-world scenario.

As an emerging paradigm for energy-efficiency design, approximate computing can reduce power consumption through simplification of logic circuits. Although calculation errors are caused by approximate computing, their impacts on the final results can be negligible in some error resilient applications, such as Convolutional Neural Networks (CNNs). Therefore, approximate computing has been applied to CNNs to reduce the high demand for computing resources and energy. Compared with the traditional method such as reducing data precision, this paper investigates the effect of approximate computing on the accuracy and power consumption of CNNs. To optimize the approximate computing technology applied to CNNs, we propose a method for quantifying the error resilience of each neuron by theoretical analysis and observe that error resilience varies widely across different neurons. On the basic of quantitative error resilience, dynamic adaptation of approximate bit-width and the corresponding configurable adder are proposed to fully exploit the error resilience of CNNs. Experimental results show that the proposed method further improves the performance of power consumption while maintaining high accuracy. By adopting the optimal approximate bit-width for each layer found by our proposed algorithm, dynamic adaptation of approximate bit-width reduces power consumption by more than 30% and causes less than 1% loss of the accuracy for LeNet-5.

The answer selection task is one of the most important issues within the automatic question answering system, and it aims to automatically find accurate answers to questions. Traditional methods for this task use manually generated features based on tf-idf and n-gram models to represent texts, and then select the right answers according to the similarity between the representations of questions and the candidate answers. Nowadays, many question answering systems adopt deep neural networks such as convolutional neural network (CNN) to generate the text features automatically, and obtained better performance than traditional methods. CNN can extract consecutive n-gram features with fixed length by sliding fixed-length convolutional kernels over the whole word sequence. However, due to the complex semantic compositionality of the natural language, there are many phrases with variable lengths and be composed of non-consecutive words in natural language, such as these phrases whose constituents are separated by other words within the same sentences. But the traditional CNN is unable to extract the variable length n-gram features and non-consecutive n-gram features. In this paper, we propose a multi-scale deformable convolutional neural network to capture the non-consecutive n-gram features by adding offset to the convolutional kernel, and also propose to stack multiple deformable convolutional layers to mine multi-scale n-gram features by the means of generating longer n-gram in higher layer. Furthermore, we apply the proposed model into the task of answer selection. Experimental results on public dataset demonstrate the effectiveness of our proposed model in answer selection.

While recent studies have shed light on the expressivity, complexity and compositionality of convolutional networks, the real inductive bias of the family of functions reachable by gradient descent on natural data is still unknown. By exploiting symmetries in the preactivation space of convolutional layers, we present preliminary empirical evidence of regularities in the preimage of trained rectifier networks, in terms of arrangements of polytopes, and relate it to the nonlinear transformations applied by the network to its input.

High frequency of network security incidents has also brought a lot of negative effects and even huge economic losses to countries, enterprises and individuals in recent years. Therefore, more and more attention has been paid to the problem of network security. In order to evaluate the newly included vulnerability text information accurately, and to reduce the workload of experts and the false negative rate of the traditional method. Multiple deep learning methods for vulnerability text classification evaluation are proposed in this paper. The standard Cross Site Scripting (XSS) vulnerability text data is processed first, and then classified using three kinds of deep neural networks (CNN, LSTM, TextRCNN) and one kind of traditional machine learning method (XGBoost). The dropout ratio of the optimal CNN network, the epoch of all deep neural networks and training set data were tuned via experiments to improve the fit on our target task. The results show that the deep learning methods evaluate vulnerability risk levels better, compared with traditional machine learning methods, but cost more time. We train our models in various training sets and test with the same testing set. The performance and utility of recurrent convolutional neural networks (TextRCNN) is highest in comparison to all other methods, which classification accuracy rate is 93.95%.

In the deep learning tasks, we can design different network models to address different tasks (classification, detection, segmentation). But traditional deep learning networks simply increase the depth and breadth of the network. This leads to a higher complexity of the model. We propose Adaptively Weighted Channel Feature Network of Mixed Convolution Kernel(SKENet). SKENet extract features from different kernels, then mixed those features by elementwise, lastly do sigmoid operator on channel features to get adaptive weightings. We did a simple classification test on the CIFAR10 amd CIFAR100 dataset. The results show that SKENet can achieve a better result in a shorter time. After that, we did an object detection experiment on the VOC dataset. The experimental results show that SKENet is far ahead of the SKNet[20] in terms of speed and accuracy.

The accuracy of the image classification by the convolutional neural network is exceeding the ability of human being and contributes to various fields. However, the improvement of the image recognition technology gives a great blow to security system with an image such as CAPTCHA. In particular, since the character string CAPTCHA has already added distortion and noise in order not to be read by the computer, it becomes a problem that the human readability is lowered. Adversarial examples is a technique to produce an image letting an image classification by the machine learning be wrong intentionally. The best feature of this technique is that when human beings compare the original image with the adversarial examples, they cannot understand the difference on appearance. However, Adversarial examples that is created with conventional FGSM cannot completely misclassify strong nonlinear networks like CNN. Osadchy et al. have researched to apply this adversarial examples to CAPTCHA and attempted to let CNN misclassify them. However, they could not let CNN misclassify character images. In this research, we propose a method to apply FGSM to the character string CAPTCHAs and to let CNN misclassified them.

With the development of the Internet, the captcha technology has also been widely used. Captcha technology is used to distinguish between humans and machines, namely Completely Automated Public Turing test to tell Computers and Humans Apart. In this paper, an end-to-end deep CNN-RNN network model is constructed by studying the captcha recognition technology, which realizes the recognition of 4-character text captcha. The CNN-RNN model first constructs a deep residual convolutional neural network based on the residual network structure to accurately extract the input captcha picture features. Then, through the constructed variant RNN network, that is, the two-layer GRU network, the deep internal features of the captcha are extracted, and finally, the output sequence is the 4-character captcha. The experiments results show that the end-to-end deep CNN-RNN network model has a good performance on different captcha datasets, achieving 99% accuracy. And experiment on the few samples dataset which only has 4000 training samples also shows an accuracy of 72.9 % and a certain generalization ability.

Frauds are known to be dynamic and have no patterns, hence they are not easy to identify. Fraudsters use recent technological advancements to their advantage. They somehow bypass security checks, leading to the loss of millions of dollars. Analyzing and detecting unusual activities using data mining techniques is one way of tracing fraudulent transactions. transactions. This paper aims to benchmark multiple machine learning methods such as k-nearest neighbor (KNN), random forest and support vector machines (SVM), while the deep learning methods such as autoencoders, convolutional neural networks (CNN), restricted boltzmann machine (RBM) and deep belief networks (DBN). The datasets which will be used are the European (EU) Australian and German dataset. The Area Under the ROC Curve (AUC), Matthews Correlation Coefficient (MCC) and Cost of failure are the 3-evaluation metrics that would be used.

Network communication data are high-dimensional and spatiotemporal, and their information content is often degraded by common traffic analysis methods. For long-term network attack detection based on network flows, it is important to extract a discriminative, high-dimensional intrinsic representation of such flows. This work focuses on a hybrid deep neural network design using a combination of a convolutional neural network (CNN) and long short-term memory (LSTM) with graph similarity measures to learn high-dimensional representations from the network traffic. In particular, examining a set of network flows, we commence by constructing a temporal communication graph and then computing graph kernel matrices. Having obtained the kernel matrices, for each graph, we use the kernel value between graphs and calculate graph characterization vectors by graph signal processing. This vector can be regarded as a kernel-based similarity embedding vector of the graph that integrates structural similarity information and leverages efficient graph kernel using the graph Laplacian matrix. Our approach exploits graph structures as the additional prior information, the graph Laplacian matrix for feature extraction and hybrid deep learning models for long-term information learning on communication graphs. Experiments on two real-world network attack datasets show that our approach can extract more discriminative representations, leading to an improved accuracy in a supervised classification task. The experimental results show that our method increases the overall accuracy by approximately 10%-15%.

This paper investigates the physical layer (PHY-layer) authentication that exploits channel state information (CSI) to enhance multiple-input multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) system security by detecting spoofing attacks in wireless networks. A multi-user authentication system is proposed using convolutional neural networks (CNNs) which also can distinguish spoofers effectively. In addition, the mini batch scheme is used to train the neural networks and accelerate the training speed. Meanwhile, L1 regularization is adopted to prevent over-fitting and improve the authentication accuracy. The convolutional-neural-network-based (CNN-based) approach can authenticate legitimate users and detect attackers by CSIs with higher performances comparing to traditional hypothesis test based methods.

Deep neural network (DNN) as a popular machine learning model is found to be vulnerable to adversarial attack. This attack constructs adversarial examples by adding small perturbations to the raw input, while appearing unmodified to human eyes but will be misclassified by a well-trained classifier. In this paper, we focus on the black-box attack setting where attackers have almost no access to the underlying models. To conduct black-box attack, a popular approach aims to train a substitute model based on the information queried from the target DNN. The substitute model can then be attacked using existing white-box attack approaches, and the generated adversarial examples will be used to attack the target DNN. Despite its encouraging results, this approach suffers from poor query efficiency, i.e., attackers usually needs to query a huge amount of times to collect enough information for training an accurate substitute model. To this end, we first utilize state-of-the-art white-box attack methods to generate samples for querying, and then introduce an active learning strategy to significantly reduce the number of queries needed. Besides, we also propose a diversity criterion to avoid the sampling bias. Our extensive experimental results on MNIST and CIFAR-10 show that the proposed method can reduce more than 90% of queries while preserve attacking success rates and obtain an accurate substitute model which is more than 85% similar with the target oracle.

Safety is one of basic human needs so we need a security system that able to prevent crime happens. Commonly, we use surveillance video to watch environment and human behaviour in a location. However, the surveillance video can only used to record images or videos with no additional information. Therefore we need more advanced camera to get another additional information such as human position and movement. This research were able to extract those information from surveillance video footage by using human detection and tracking algorithm. The human detection framework is based on Deep Learning Convolutional Neural Networks which is a very popular branch of artificial intelligence. For tracking algorithms, channel and spatial correlation filter is used to track detected human. This system will generate and export tracked movement on footage as an additional information. This tracked movement can be analysed furthermore for another research on surveillance video problems.

Video retrieval technology faces a series of challenges with the tremendous growth in the number of videos. In order to improve the retrieval performance in efficiency and accuracy, a novel deep hash method for video data hierarchical retrieval is proposed in this paper. The approach first uses cluster-based method to extract key frames, which reduces the workload of subsequent work. On the basis of this, high-level semantical features are extracted from VGG16, a widely used deep convolutional neural network (deep CNN) model. Then we utilize a hierarchical retrieval strategy to improve the retrieval performance, roughly can be categorized as coarse search and fine search. In coarse search, we modify simHash to learn hash codes for faster speed, and in fine search, we use the Euclidean distance to achieve higher accuracy. Finally, we compare our approach with other two methods through practical experiments on two videos, and the results demonstrate that our approach has better retrieval effect.

Railroad trespassing is a dangerous activity with significant security and safety risks. However, regular patrolling of potential trespassing sites is infeasible due to exceedingly high resource demands and personnel costs. This raises the need to design automated trespass detection and early warning prediction techniques leveraging state-of-the-art machine learning. To meet this need, we propose a novel framework for Automated Railroad Trespassing detection System using video surveillance data called ARTS. As the core of our solution, we adopt a CNN-based deep learning architecture capable of video processing. However, these deep learning-based methods, while effective, are known to be computationally expensive and time consuming, especially when applied to a large volume of surveillance data. Leveraging the sparsity of railroad trespassing activity, ARTS corresponds to a dual-stage deep learning architecture composed of an inexpensive pre-filtering stage for activity detection, followed by a high fidelity trespass classification stage employing deep neural network. The resulting dual-stage ARTS architecture represents a flexible solution capable of trading-off accuracy with computational time. We demonstrate the efficacy of our approach on public domain surveillance data achieving 0.87 f1 score while keeping up with the enormous video volume, achieving a practical time and accuracy trade-off.

The progressed computational abilities of numerous asset compelled gadgets mobile phones have empowered different research zones including picture recovery from enormous information stores for various IoT applications. The real difficulties for picture recovery utilizing cell phones in an IoT situation are the computational intricacy and capacity. To manage enormous information in IoT condition for picture recovery a light-weighted profound learning base framework for vitality obliged gadgets. The framework initially recognizes and crop face areas from a picture utilizing Viola-Jones calculation with extra face classifier to take out the identification issue. Besides, the utilizes convolutional framework layers of a financially savvy pre-prepared CNN demonstrate with characterized highlights to speak to faces. Next, highlights of the huge information vault are listed to accomplish a quicker coordinating procedure for constant recovery. At long last, Euclidean separation is utilized to discover comparability among question and archive pictures. For exploratory assessment, we made a nearby facial pictures dataset it including equally single and gathering face pictures. In the dataset can be utilized by different specialists as a scale for examination with other ongoing facial picture recovery frameworks. The trial results demonstrate that our planned framework beats other cutting edge highlight extraction strategies as far as proficiency and recovery for IoT-helped vitality obliged stages.

Anomaly detection generally involves the extraction of features from entities' or users' properties, and the design of anomaly detection models using machine learning or deep learning algorithms. However, only considering entities' property information could lead to high false positives. We posit the importance of also considering connections or relationships between entities in the detecting of anomalous behaviors and associated threat groups. Therefore, in this paper, we design a GCN (graph convolutional networks) based anomaly detection model to detect anomalous behaviors of users and malicious threat groups. The GCN model could characterize entities' properties and structural information between them into graphs. This allows the GCN based anomaly detection model to detect both anomalous behaviors of individuals and associated anomalous groups. We then evaluate the proposed model using a real-world insider threat data set. The results show that the proposed model outperforms several state-of-art baseline methods (i.e., random forest, logistic regression, SVM, and CNN). Moreover, the proposed model can also be applied to other anomaly detection applications.

Recent breakthrough in neural network has led to the birth of Convolutional neural network (CNN) which has been found to be very efficient especially in the areas of image recognition and classification. This success is traceable to the availability of large datasets and its capability to learn salient and complex data features which subsequently produce a reusable output model (Fθ). The Fθ are often made available (e.g. on cloud as-a-service) for others (client) to train their data or do transfer learning, however, an adversary can perpetrate a model inversion attack on the model Fθ to recover training data, hence compromising the sensitivity of the model buildup data. This is possible because CNN as a variant of deep neural network does memorize most of its training data during learning. Consequently, this has pose a privacy concern especially when a medical or financial data are used as model buildup data. Existing researches that proffers privacy preserving approach however suffer from significant accuracy degradation and this has left privacy preserving model on a theoretical desk. In this paper, we proposed an ϵ-tuple differential privacy approach that is based on neuron impact factor estimation to preserve privacy of CNN model without significant accuracy degradation. We experiment our approach on two large datasets and the result shows no significant accuracy degradation.

In this paper, a merged convolution neural network (MCNN) is proposed to improve the accuracy and robustness of real-time facial expression recognition (FER). Although there are many ways to improve the performance of facial expression recognition, a revamp of the training framework and image preprocessing renders better results in applications. When the camera is capturing images at high speed, however, changes in image characteristics may occur at certain moments due to the influence of light and other factors. Such changes can result in incorrect recognition of human facial expression. To solve this problem, we propose a statistical method for recognition results obtained from previous images, instead of using the current recognition output. Experimental results show that the proposed method can satisfactorily recognize seven basic facial expressions in real time.

Deep learning is a popular powerful machine learning solution to the computer vision tasks. The most criticized vulnerability of deep learning is its poor tolerance towards adversarial images obtained by deliberately adding imperceptibly small perturbations to the clean inputs. Such negatives can delude a classifier into wrong decision making. Previous defensive techniques mostly focused on refining the models or input transformation. They are either implemented only with small datasets or shown to have limited success. Furthermore, they are rarely scrutinized from the hardware perspective despite Artificial Intelligence (AI) on a chip is a roadmap for embedded intelligence everywhere. In this paper we propose a new discriminative noise injection strategy to adaptively select a few dominant layers and progressively discriminate adversarial from benign inputs. This is made possible by evaluating the differences in label change rate from both adversarial and natural images by injecting different amount of noise into the weights of individual layers in the model. The approach is evaluated on the ImageNet Dataset with 8-bit truncated models for the state-of-the-art DNN architectures. The results show a high detection rate of up to 88.00% with only approximately 5% of false positive rate for MobileNet. Both detection rate and false positive rate have been improved well above existing advanced defenses against the most practical noninvasive universal perturbation attack on deep learning based AI chip.

The storage efficiency of hash codes and their application in the fast approximate nearest neighbor search, along with the explosion in the size of available labeled image datasets caused an intensive interest in developing learning based hash algorithms recently. In this paper, we present a learning based hash algorithm that utilize ordinal information of feature vectors. We have proposed a novel mathematically differentiable approximation of argmax function for this hash algorithm. It has enabled seamless integration of hash function with deep neural network architecture which can exploit the rich feature vectors generated by convolutional neural networks. We have also proposed a loss function for the case that the hash code is not binary and its entries are digits of arbitrary k-ary base. The resultant model comprised of feature vector generation and hashing layer is amenable to end-to-end training using gradient descent methods. In contrast to the majority of current hashing algorithms that are either not learning based or use hand-crafted feature vectors as input, simultaneous training of the components of our system results in better optimization. Extensive evaluations on NUS-WIDE, CIFAR-10 and MIRFlickr benchmarks show that the proposed algorithm outperforms state-of-art and classical data agnostic, unsupervised and supervised hashing methods by 2.6% to 19.8% mean average precision under various settings.

In this paper, we present an extensive evaluation of face recognition and verification approaches performed by the European COST Action MULTI-modal Imaging of FOREnsic SciEnce Evidence (MULTI-FORESEE). The aim of the study is to evaluate various face recognition and verification methods, ranging from methods based on facial landmarks to state-of-the-art off-the-shelf pre-trained Convolutional Neural Networks (CNN), as well as CNN models directly trained for the task at hand. To fulfill this objective, we carefully designed and implemented a realistic data acquisition process, that corresponds to a typical face verification setup, and collected a challenging dataset to evaluate the real world performance of the aforementioned methods. Apart from verifying the effectiveness of deep learning approaches in a specific scenario, several important limitations are identified and discussed through the paper, providing valuable insight for future research directions in the field.

In this paper, an intruder detection system has been built with an implementation of convolutional neural network (CNN) using raspberry pi, Microsoft's Azure and Twilio cloud systems. The CNN algorithm which is stored in the cloud is implemented to basically classify input data as either intruder or user. By using the raspberry pi as the middleware and raspberry pi camera for image acquisition, efficient execution of the learning and classification operations are performed using higher resources that cloud computing offers. The cloud system is also programmed to alert designated users via multimedia messaging services (MMS) when intruders or users are detected. Furthermore, our work has demonstrated that, though convolutional neural network could impose high computing demands on a processor, the input data could be obtained with low-cost modules and middleware which are of low processing power while subjecting the actual learning algorithm execution to the cloud system.