Biblio

Facial emotion recognition helps feed the growth of the future artificial intelligence with the development of emotion recognition, learning, and analysis of different angles of a human face and head pose. The world's recent pandemic gave rise to the rapid installment of facial recognition for fewer applications, while emotion recognition is still within the experimental boundaries. The current challenges encountered with facial emotion recognition (FER) are the difference between background noises. Since today's world shows us that humans soon need robotics in the most significant role of human perception, attention, memory, decision-making, and human-robot interaction (HRI) needs employees. By merging the head pose as a combination towards the FER to boost the robustness in understanding emotions using the convolutional neural networks (CNN). The stochastic gradient descent with a comprehensive model is adopted by applying multi-task learning capable of implicit parallelism, inherent and better global optimizer in finding better network weights. After executing a multi-task learning model using two independent datasets, the experiment with the FER and head pose learning multi-views co-regularization frameworks were subsequently merged with validation accuracy.

Child facial expression recognition is a relatively less investigated area within affective computing. Children’s facial expressions differ significantly from adults; thus, it is necessary to develop emotion recognition frameworks that are more objective, descriptive and specific to this target user group. In this paper we propose the first approach that (i) constructs video-level heterogeneous graph representation for facial expression recognition in children, and (ii) predicts children’s facial expressions using the automatically detected Action Units (AUs). To this aim, we construct three separate length-independent representations, namely, statistical, spectral and graph at video-level for detailed multi-level facial behaviour decoding (AU activation status, AU temporal dynamics and spatio-temporal AU activation patterns, respectively). Our experimental results on the LIRIS Children Spontaneous Facial Expression Video Database demonstrate that combining these three feature representations provides the highest accuracy for expression recognition in children.

Recently deep learning has demonstrated much success within the fields of image and natural language processing, facial recognition, and computer vision. The success is attributed to large, accessible databases and deep learning's ability to learn highly accurate models. Thus, deep learning is being investigated as a viable end-to-end approach to digital communications design. This work investigates the use of adversarial deep learning to ensure that a radio can communicate covertly, via Direct Sequence Spread Spectrum (DSSS), with another while a third (the adversary) is actively attempting to detect, intercept and exploit their communications. The adversary's ability to detect and exploit the DSSS signals is hindered by: (i) generating a set of spreading codes that are balanced and result in low side lobes as well as (ii) actively adapting the encoding scheme. Lastly, DSSS communications performance is assessed using energy constrained devices to accurately portray IoT and IoBT device limitations.

In this paper, our aim is to detect illegal spectrum access behaviors. Firstly, we detect whether the channel is busy, and then if it is busy, recognizing whether there are illegal users. To get closer to the actual situation, we consider a more general scenario where multiple users are authorized to work on the same channel under certain interference control strategies, and build it as a ternary hypothesis test model using the generalized multi-hypothesis Neyman-Pearson criterion. Considering the various potential combination of multiple authorized users, the spectrum detection process utilizes a two-step detector. We adopt the Generalized Likelihood Ratio Test (GLRT) and the Rao test to detect illegal spectrum access behaviors. What is more, the Wald test is proposed which has a compromise between computational complexity and performance. The relevant formulas of the three detection schemes are derived. Finally, comprehensive and in-depth simulations are provided to verify the effectiveness of the proposed detection scheme that it has the best detection performance under different authorized sample numbers and different performance constraints. Besides, we illustrate the probability of detection of illegal behaviors under different parameters of illegal behaviors and different sets of AUs' states under the Wald test.

In order to solve the problem of accurate positioning of mixed reality virtual scene in physical space, this paper, firstly, analyzes the positioning principle of mixed reality virtual scene. Secondly, based on the comparison among the three developer kits: ARToolKit, ARTag, and Vuforia and two image optimization algorithms: AHE and ACE, it makes sure to use Vuforia development tool to complete the signature-based tracking and registration task, and use ACE algorithm to optimize the signature-based image. It improves the efficiency, stability and accuracy of image recognition registration. Then the multi-target recognition and registration technology is used to realize the multi-location of virtual scene. Finally, Hololens glasses are used as the hardware carrier to verify the above method. The experimental results show that the above method not only realizes the precise location of MR virtual scene based on image recognition, but also ensures the absolute position of the virtual model in the real space, bringing users a more real virtual experience. Keywords-mixed reality, multi-person collaboration, virtual positioning, gesture interaction.

Recommender systems (RSs) are intelligent systems to help e-commerce users to find their preferred items among millions of available items by considering the profiles of both users and items. These systems need to predict the unknown ratings and then recommend a set of high rated items. Among the others, Collaborative Filtering (CF) is a successful recommendation approach and has been utilized in many real-world systems. CF methods seek to predict missing ratings by considering the preferences of those users who are similar to the target user. A major task in Collaborative Filtering is to identify an accurate set of users and employing them in the rating prediction process. Most of the CF-based methods suffer from the cold-start issue which arising from an insufficient number of ratings in the prediction process. This is due to the fact that users only comment on a few items and thus CF methods faced with a sparse user-item matrix. To tackle this issue, a new collaborative filtering method is proposed that has a trust-aware strategy. The proposed method employs the trust relationships of users as additional information to help the CF tackle the cold-start issue. To this end, the proposed integrated trust relationships in the prediction process by using the Ant Colony Optimization (ACO). The proposed method has four main steps. The aim of the first step is ranking users based on their similarities to the target user. This step uses trust relationships and the available rating values in its process. Then in the second step, graph clustering methods are used to cluster the trust graph to group similar users. In the third step, the users are weighted based on their similarities to the target users. To this end, an ACO process is employed on the users' graph. Finally, those of top users with high similarity to the target user are used in the rating prediction process. The superiority of our method has been shown in the experimental results in comparison with well-known and state-of-the-art methods.

Modern learning systems, say a tutoring platform, has many characteristics like digital data presentation with interactivity, mobility, which provides information about the study-content as per the learners understanding levels, intelligent learners behavior, etc. A sophisticated ubiquitous learner system maintains security and monitors the mischievous behavior of the learner, and authenticates and authorizes every learner, which is quintessential. Some of the existing security schemes aim only at single entry-point authentication, which may not suit to ubiquitous tutor platform. We propose a secured authentication scheme which is based on the information utility of the learner. Whenever a learner moves into a tutor platform, which has ubiquitous learner system technology, the system at first-begins with learners' identity authentication, and then it initiates trust evaluation after the successful authentication of the learner. Periodic credential verification of the learner will be carried out, which intensifies the authentication scheme of the system proposed. BAN logic has been used to prove the authentication in this system. The proposed authentication scheme has been simulated and analyzed for the indoor tutor platform environment.

When disasters occur, they not only affect the human life. Therefore, communication in disaster management is very important. During the disaster recovery phase, the network infrastructure may be partially fragmented and mobile rescue operations may involve many teams with different roles which can dynamically change. Therefore, disaster management services require high flexibility both in terms of network infrastructure management and rescue group communication. Existing studies have shown that IP-based or traditional telephony solutions are not well-suited to deal with such flexible group communication and network management due to their connection-oriented communication, no built-in support for mobile devices, and no mechanism for network fragmentation. Recent studies show that information-centric networking offers scalable and flexible communication based on its name-based interest-oriented communication approach. However, considering the difficulty of deploying a new service on the existing network, the programmability and virtualization of the network are required. This paper presents our implementation of an information-centric disaster management system based on network function virtualization (vICSNF). We show a proof-of-concept system with a case study for Seoul disaster management services. The system achieves flexibility both in terms of network infrastructure management and rescue group communication. Obtained testbed results show that vICSNF achieves a low communication overhead compared to the IP-based approach and the auto-configuration of vICSNFs enables the quick deployment for disaster management services in disaster scenarios.

This paper deals with novel group-based Authentication and Key Agreement protocol for Internet of Things(IoT) enabled LTE/LTE-A network to overcome the problems of computational overhead, complexity and problem of heterogeneous devices, where other existing methods are lagging behind in attaining security requirements and computational overhead. In this work, two Groups are created among Machine Type Communication Devices (MTCDs) on the basis of device type to reduce complexity and problems of heterogeneous devices. This paper fulfills all the security requirements such as preservation, mutual authentication, confidentiality. Bio-metric authentication has been used to enhance security level of the network. The security and performance analysis have been verified through simulation results. Moreover, the performance of the proposed Novel Group-Based Authentication and key Agreement(AKA) Protocol is analyzed with other existing IoT enabled LTE/LTE-A protocol.

We present a fast dynamic graph data structure for the GPU. Our dynamic graph structure uses one hash table per vertex to store adjacency lists and achieves 3.4-14.8x faster insertion rates over the state of the art across a diverse set of large datasets, as well as deletion speedups up to 7.8x. The data structure supports queries and dynamic updates through both edge and vertex insertion and deletion. In addition, we define a comprehensive evaluation strategy based on operations, workloads, and applications that we believe better characterize and evaluate dynamic graph data structures.

Collective self-adaptive systems (CSAS) are distributed and interconnected systems composed of multiple agents that can perform complex tasks such as environmental data collection, search and rescue operations, and discovery of natural resources. By providing individual agents with learning capabilities, CSAS can cope with challenges related to distributed sensing and decision-making and operate in uncertain environments. This unique characteristic of CSAS enables the collective to exhibit robust behaviour while achieving system-wide and agent-specific goals. Although learning has been explored in many CSAS applications, selecting suitable learning models and techniques remains a significant challenge that is heavily influenced by expert knowledge. We address this gap by performing a multifaceted analysis of existing CSAS with learning capabilities reported in the literature. Based on this analysis, we introduce a 3D framework that illustrates the learning aspects of CSAS considering the dimensions of autonomy, knowledge access, and behaviour, and facilitates the selection of learning techniques and models. Finally, using example applications from this analysis, we derive open challenges and highlight the need for research on collaborative, resilient and privacy-aware mechanisms for CSAS.

As a new generation of network architecture with subversive changes to traditional IP networks, Content Centric Networks (CCN) has attracted widespread attention from domestic and foreign scholars for its efficient content distribution, multi-path and secure routing features. The design architecture of CCN network has many advantages. However, it is also easily used illegally, which brings certain security problems. For example, objectified network resources which include requesters, publishers, content and node routes, are faced with many security threats, such as privacy attribute disclosure, privacy detection, content information disclosure, and spoofing and denial of service attacks. A node routing security scheme based on dynamic reputation value is proposed for the security problem of node routing. It is convenient for detecting node routing attacks and defending in time. And it could provide security for the Content Centric Networks node routing without affecting the node routing advantages and normal user requests.

Wireless technology has seen a tremendous growth in the recent past. Orthogonal Frequency Division Multiplexing (OFDM) modulation scheme has been utilized in almost all the advanced wireless techniques because of the advantages it offers. Hence in this aspect, SystemVue based OFDM transceiver has been developed with AWGN as the channel noise. To mitigate the channel noise Convolutional code with Viterbi decoder has been depicted. Further to protect the information from the malicious users the data is scrambled with the aid of gold codes. The performance of the transceiver is analysed through various Bit Error Rate (BER) versus Signal to Noise Ratio (SNR) graphs.

Platoon is one of cooperative driving applications where a set of vehicles can collaboratively sense each other for driving safety and traffic efficiency. However, platoon without security insurance makes the cooperative vehicles vulnerable to cyber-attacks, which may cause life-threatening accidents. In this paper, we introduce malicious attacks in platoon maneuvers. To defend against these attacks, we propose a Cyphertext-Policy Attribute-Based Encryption (CP-ABE) based Platoon Secure Sensing scheme, named CPSS. In the CPSS, platoon key is encapsulated in the access control structure in the key distribution process, so that interference messages sending by attackers without the platoon key could be ignored. Therefore, the sensing data which contains speed and position information can be protected. In this way, speed and distance fluctuations caused by attacks can be mitigated even eliminated thereby avoiding the collisions and ensuring the overall platoon stability. Time complexity analysis shows that the CPSS is more efficient than that of the polynomial time solutions. Finally, to evaluate capabilities of the CPSS, we integrate a LTE-V2X with platoon maneuvers based on Veins platform. The evaluation results show that the CPSS outperforms the baseline algorithm by 25% in terms of distance variations.

Social media has been one of the most efficacious and precise by speakers of public opinion. A strategy which sanctions the utilization and illustration of twitter data to conclude public conviction is discussed in this paper. Sentiments on exclusive entities with diverse strengths and intenseness are stated by public, where these sentiments are strenuously cognate to their personal mood and emotions. To examine the sentiments from natural language texts, addressing various opinions, a lot of methods and lexical resources have been propounded. A path for boosting twitter sentiment classification using various sentiment proportions as meta-level features has been proposed by this article. Analysis of tweets was done on the product iPhone 6.

The past decade has seen a growing interest in underwater acoustic positioning networks (UAPNs) because of their wide applications in marine research, ocean monitoring, offshore exploration, and defense or homeland security. Efficient communication among all sensors and receivers is crucial so as to make positioning service available. Traditional UAPNs could locate only one target, that are growing obsolete due to increasing demands for multiple users working at the same time. Due to the demands for multiple users working simultaneously and narrow acoustic bandwidth, new efficient and reliable communication and networking protocols are required in design for UAPNs. In this paper, we aim to provide the procedure of communication design for UAPNs based on sonar equation and spread spectrum communication. What's more, signal design and performance analysis are supplied. The results show that the signal we designed have ideal correlation performance and high processing gain. The signal is suitable for multiple users UAPNs and thus show favorable potential in ocean engineering applications.

Facial expression is the most powerful and natural non-verbal emotional communication method. Facial Expression Recognition(FER) has significance in machine learning tasks. Deep Learning models perform well in FER tasks, but it doesn't provide any justification for its decisions. Based on the hypothesis that facial expression is a combination of facial muscle movements, we find that Facial Action Coding Units(AUs) and Emotion label have a relationship in CK+ Dataset. In this paper, we propose a model which utilises AUs to explain Convolutional Neural Network(CNN) model's classification results. The CNN model is trained with CK+ Dataset and classifies emotion based on extracted features. Explanation model classifies the multiple AUs with the extracted features and emotion classes from the CNN model. Our experiment shows that with only features and emotion classes obtained from the CNN model, Explanation model generates AUs very well.

Computing the probability of vulnerability exploitation in Bayesian attack graphs (BAGs) is a key process for the network security assessment. The conditional probability of vulnerability exploitation could be obtained from the exploitability of the NIST's Common Vulnerability Scoring System (CVSS). However, the method which N. Poolsappasit et al. proposed for computing conditional probability could be used only in the CVSS metric version v2.0, and can't be used in other two versions. In this paper, we present two methods for computing the conditional probability based on CVSS's other two metric versions, version 1.0 and version 3.0, respectively. Based on the CVSS, the conditional probability computation of vulnerability exploitation is complete by combining the method of N. Poolsappasit et al.

Recently, Jung et al. [1] proposed a data access privilege scheme and claimed that their scheme addresses data and identity privacy as well as multi-authority, and provides data access privilege for attribute-based encryption. In this paper, we show that this scheme, and also its former and latest versions (i.e. [2] and [3] respectively) suffer from a number of weaknesses in terms of finegrained access control, users and authorities collusion attack, user authorization, and user anonymity protection. We then propose our new scheme that overcomes these shortcomings. We also prove the security of our scheme against user collusion attacks, authority collusion attacks and chosen plaintext attacks. Lastly, we show that the efficiency of our scheme is comparable with existing related schemes.

Security of control systems have become a new and important field of research since malicious attacks on control systems indeed occurred including Stuxnet in 2011 and north eastern electrical grid black out in 2003. Attacks on sensors and/or actuators of control systems cause malfunction, instability, and even system destruction. The impact of attack may differ by which instrumentation (sensors and/or actuators) is being attacked. In particular, for control systems with multiple sensors, attack on each sensor may have different impact, i.e., attack on some sensors leads to a greater damage to the system than those for other sensors. To investigate this, we consider sensor bias injection attacks in linear control systems equipped with anomaly detector, and quantify the maximum impact of attack on sensors while the attack remains undetected. Then, we introduce a notion of sensor security index for linear dynamic systems to quantify the vulnerability under sensor attacks. Method of reducing system vulnerability is also discussed using the notion of sensor security index.

Strecth Processing (SP) is a radar signal processing technique that provides high-range resolution with processing large bandwidth signals with lower rate Analog to Digital Converter(ADC)s. The range resolution of the large bandwidth signal is obtained through looking into a limited range window and low rate ADC samples. The target space in the observed range window is sparse and Compressive sensing(CS) is an important tool to further decrease the number of measurements and sparsely reconstruct the target space for sparse scenes with a known basis which is the Fourier basis in the general application of SP. Although classical CS techniques might be directly applied to SP, due to off-grid targets reconstruction performance degrades. In this paper, applicability of compressive sensing framework and its sparse signal recovery techniques to stretch processing is studied considering off-grid cases. For sparsity based robust SP, Perturbed Parameter Orthogonal Matching Pursuit(PPOMP) algorithm is proposed. PPOMP is an iterative technique that estimates off-grid target parameters through a gradient descent. To compute the error between actual and reconstructed parameters, Earth Movers Distance(EMD) is used. Performance of proposed algorithm are compared with classical CS and SP techniques.

Multi-touch attribution, which allows distributing the credit to all related advertisements based on their corresponding contributions, has recently become an important research topic in digital advertising. Traditionally, rule-based attribution models have been used in practice. The drawback of such rule-based models lies in the fact that the rules are not derived form the data but only based on simple intuition. With the ever enhanced capability to tracking advertisement and users' interaction with the advertisement, data-driven multi-touch attribution models, which attempt to infer the contribution from user interaction data, become an important research direction. We here propose a new data-driven attribution model based on survival theory. By adopting a probabilistic framework, one key advantage of the proposed model is that it is able to remove the presentation biases inherit to most of the other attribution models. In addition to model the attribution, the proposed model is also able to predict user's 'conversion' probability. We validate the proposed method with a real-world data set obtained from a operational commercial advertising monitoring company. Experiment results have shown that the proposed method is quite promising in both conversion prediction and attribution.

The trusted network connection is a hot spot in trusted computing field and the trust measurement and access control technology are used to deal with network security threats in trusted network. But the trusted network connection lacks fine-grained states and real-time measurement support for the client and the authentication mechanism is difficult to apply in the trusted network connection, it is easy to cause the loss of identity privacy. In order to solve the above-described problems, this paper presents a trust measurement scheme suitable for clients in the trusted network, the scheme integrates the following attributes such as authentication mechanism, state measurement, and real-time state measurement and so on, and based on the authentication mechanism and the initial state measurement, the scheme uses the real-time state measurement as the core method to complete the trust measurement for the client. This scheme presented in this paper supports both static and dynamic measurements. Overall, the characteristics of this scheme such as fine granularity, dynamic, real-time state measurement make it possible to make more fine-grained security policy and therefore it overcomes inadequacies existing in the current trusted network connection.