Visible to the public Biblio

Found 145 results

Filters: First Letter Of Last Name is I  [Clear All Filters]
A B C D E F G H [I] J K L M N O P Q R S T U V W X Y Z   [Show ALL]
I
Iordanou, Costas, Smaragdakis, Georgios, Poese, Ingmar, Laoutaris, Nikolaos.  2018.  Tracing Cross Border Web Tracking. Proceedings of the Internet Measurement Conference 2018. :329-342.

A tracking flow is a flow between an end user and a Web tracking service. We develop an extensive measurement methodology for quantifying at scale the amount of tracking flows that cross data protection borders, be it national or international, such as the EU28 border within which the General Data Protection Regulation (GDPR) applies. Our methodology uses a browser extension to fully render advertising and tracking code, various lists and heuristics to extract well known trackers, passive DNS replication to get all the IP ranges of trackers, and state-of-the art geolocation. We employ our methodology on a dataset from 350 real users of the browser extension over a period of more than four months, and then generalize our results by analyzing billions of web tracking flows from more than 60 million broadband and mobile users from 4 large European ISPs. We show that the majority of tracking flows cross national borders in Europe but, unlike popular belief, are pretty well confined within the larger GDPR jurisdiction. Simple DNS redirection and PoP mirroring can increase national confinement while sealing almost all tracking flows within Europe. Last, we show that cross boarder tracking is prevalent even in sensitive and hence protected data categories and groups including health, sexual orientation, minors, and others.

Iorga, Denis, Corlătescu, Dragos, Grigorescu, Octavian, Săndescu, Cristian, Dascălu, Mihai, Rughiniş, Razvan.  2020.  Early Detection of Vulnerabilities from News Websites using Machine Learning Models. 2020 19th RoEduNet Conference: Networking in Education and Research (RoEduNet). :1–6.
The drawbacks of traditional methods of cybernetic vulnerability detection relate to the required time to identify new threats, to register them in the Common Vulnerabilities and Exposures (CVE) records, and to score them with the Common Vulnerabilities Scoring System (CVSS). These problems can be mitigated by early vulnerability detection systems relying on social media and open-source data. This paper presents a model that aims to identify emerging cybernetic vulnerabilities in cybersecurity news articles, as part of a system for automatic detection of early cybernetic threats using Open Source Intelligence (OSINT). Three machine learning models were trained on a novel dataset of 1000 labeled news articles to create a strong baseline for classifying cybersecurity articles as relevant (i.e., introducing new security threats), or irrelevant: Support Vector Machines, a Multinomial Naïve Bayes classifier, and a finetuned BERT model. The BERT model obtained the best performance with a mean accuracy of 88.45% on the test dataset. Our experiments support the conclusion that Natural Language Processing (NLP) models are an appropriate choice for early vulnerability detection systems in order to extract relevant information from cybersecurity news articles.
Iosifidis, Efthymios, Limniotis, Konstantinos.  2016.  A Study of Lightweight Block Ciphers in TLS: The Case of Speck. Proceedings of the 20th Pan-Hellenic Conference on Informatics. :64:1–64:5.

Application of lightweight block ciphers in the TLS protocol is studied in this paper. More precisely, since the use of lightweight cryptographic algorithms is prerequisite for addressing security in highly constrained environments such as the Internet of Things, we focus on the behavior of the TLS performance in case that AES is being replaced by a lightweight block cipher; to this end, the recently proposed Speck cipher is being used as a case study. Experimental results exhibit that significant gain in performance can be achieved in such constrained environments, whereas in some cases Speck with larger key size than AES may also result in higher throughput.

Ippisch, A., Graffi, K..  2017.  Infrastructure Mode Based Opportunistic Networks on Android Devices. 2017 IEEE 31st International Conference on Advanced Information Networking and Applications (AINA). :454–461.

Opportunistic Networks are delay-tolerant mobile networks with intermittent node contacts in which data is transferred with the store-carry-forward principle. Owners of smartphones and smart objects form such networks due to their social behaviour. Opportunistic Networking can be used in remote areas with no access to the Internet, to establish communication after disasters, in emergency situations or to bypass censorship, but also in parallel to familiar networking. In this work, we create a mobile network application that connects Android devices over Wi-Fi, offers identification and encryption, and gathers information for routing in the network. The network application is constructed in such a way that third party applications can use the network application as network layer to send and receive data packets. We create secure and reliable connections while maintaining a high transmission speed, and with the gathered information about the network we offer knowledge for state of the art routing protocols. We conduct tests on connectivity, transmission range and speed, battery life and encryption speed and show a proof of concept for routing in the network.

Iqbal, A., Mahmood, F., Shalaginov, A., Ekstedt, M..  2018.  Identification of Attack-based Digital Forensic Evidences for WAMPAC Systems. 2018 IEEE International Conference on Big Data (Big Data). :3079–3087.
Power systems domain has generally been very conservative in terms of conducting digital forensic investigations, especially so since the advent of smart grids. This lack of research due to a multitude of challenges has resulted in absence of knowledge base and resources to facilitate such an investigation. Digitalization in the form of smart grids is upon us but in case of cyber-attacks, attribution to such attacks is challenging and difficult if not impossible. In this research, we have identified digital forensic artifacts resulting from a cyber-attack on Wide Area Monitoring, Protection and Control (WAMPAC) systems, which will help an investigator attribute an attack using the identified evidences. The research also shows the usage of sandboxing for digital forensics along with hardware-in-the-loop (HIL) setup. This is first of its kind effort to identify and acquire all the digital forensic evidences for WAMPAC systems which will ultimately help in building a body of knowledge and taxonomy for power system forensics.
Iqbal, H., Ma, J., Mu, Q., Ramaswamy, V., Raymond, G., Vivanco, D., Zuena, J..  2017.  Augmenting Security of Internet-of-Things Using Programmable Network-Centric Approaches: A Position Paper. 2017 26th International Conference on Computer Communication and Networks (ICCCN). :1–6.

Advances in nanotechnology, large scale computing and communications infrastructure, coupled with recent progress in big data analytics, have enabled linking several billion devices to the Internet. These devices provide unprecedented automation, cognitive capabilities, and situational awareness. This new ecosystem–termed as the Internet-of-Things (IoT)–also provides many entry points into the network through the gadgets that connect to the Internet, making security of IoT systems a complex problem. In this position paper, we argue that in order to build a safer IoT system, we need a radically new approach to security. We propose a new security framework that draws ideas from software defined networks (SDN), and data analytics techniques; this framework provides dynamic policy enforcements on every layer of the protocol stack and can adapt quickly to a diverse set of industry use-cases that IoT deployments cater to. Our proposal does not make any assumptions on the capabilities of the devices - it can work with already deployed as well as new types of devices, while also conforming to a service-centric architecture. Even though our focus is on industrial IoT systems, the ideas presented here are applicable to IoT used in a wide array of applications. The goal of this position paper is to initiate a dialogue among standardization bodies and security experts to help raise awareness about network-centric approaches to IoT security.

Iqbal, Maryam, Iqbal, Mohammad Ayman.  2019.  Attacks Due to False Data Injection in Smart Grids: Detection Protection. 2019 1st Global Power, Energy and Communication Conference (GPECOM). :451-455.

As opposed to a traditional power grid, a smart grid can help utilities to save energy and therefore reduce the cost of operation. It also increases reliability of the system In smart grids the quality of monitoring and control can be adequately improved by incorporating computing and intelligent communication knowledge. However, this exposes the system to false data injection (FDI) attacks and the system becomes vulnerable to intrusions. Therefore, it is important to detect such false data injection attacks and provide an algorithm for the protection of system against such attacks. In this paper a comparison between three FDI detection methods has been made. An H2 control method has then been proposed to detect and control the false data injection on a 12th order model of a smart grid. Disturbances and uncertainties were added to the system and the results show the system to be fully controllable. This paper shows the implementation of a feedback controller to fully detect and mitigate the false data injection attacks. The controller can be incorporated in real life smart grid operations.

Iqbal, Shahrear, Haque, Anwar, Zulkernine, Mohammad.  2019.  Towards a Security Architecture for Protecting Connected Vehicles from Malware. 2019 IEEE 89th Vehicular Technology Conference (VTC2019-Spring). :1—5.

Vehicles are becoming increasingly connected to the outside world. We can connect our devices to the vehicle's infotainment system and internet is being added as a functionality. Therefore, security is a major concern as the attack surface has become much larger than before. Consequently, attackers are creating malware that can infect vehicles and perform life-threatening activities. For example, a malware can compromise vehicle ECUs and cause unexpected consequences. Hence, ensuring the security of connected vehicle software and networks is extremely important to gain consumer confidence and foster the growth of this emerging market. In this paper, we propose a characterization of vehicle malware and a security architecture to protect vehicle from these malware. The architecture uses multiple computational platforms and makes use of the virtualization technique to limit the attack surface. There is a real-time operating system to control critical vehicle functionalities and multiple other operating systems for non-critical functionalities (infotainment, telematics, etc.). The security architecture also describes groups of components for the operating systems to prevent malicious activities and perform policing (monitor, detect, and control). We believe this work will help automakers guard their systems against malware and provide a clear guideline for future research.

Irazoqui, Gorka, Eisenbarth, Thomas, Sunar, Berk.  2018.  MASCAT: Preventing Microarchitectural Attacks Before Distribution. Proceedings of the Eighth ACM Conference on Data and Application Security and Privacy. :377–388.
Microarchitectural attacks have gained popularity lately for the threat they pose and for their stealthiness. They are stealthy as they only exploit common harmless resources accessible at lowest privilege level, e.g. timed memory and cache accesses. Microarchitectural attacks have proven successful on shared cloud instances across VMs, on smartphones with sandboxing, and on numerous embedded platforms. Further they have shown to have catastrophic consequences such as critical data recovery or memory isolation bypassing. Due to the rise of malicious code, app store operators such as Microsoft, Apple and Google are already vetting apps before releasing them. Microarchitectural attacks however still bypass such detection mechanisms as they mainly utilize standard resources and look harmless. Given the rise of malicious code in app stores and in online repositories it becomes essential to scan applications for such stealthy attacks to prevent their distribution. We present a static code analysis tool, MASCAT, capable of scanning for ever-evolving microarchitectural attacks. MASCAT can be used by app store service providers to perform large scale fully automated analysis of applications. The initial MASCAT suite is built to include cache/DRAM access attacks and rowhammer. MASCAT detects several patterns that are common and necessary to execute microarchitectural attacks. MASCAT currently has a detection rate of 96% and an average false positive rate tested in 1200 applications of 0.75%. Further, our tool can easily be extended to cover newer attack vectors as they emerge
Iriqat, Yousef Mohammad, Ahlan, Abd Rahman, Molok, Nurul Nuha Abdul.  2019.  Information Security Policy Perceived Compliance Among Staff in Palestine Universities: An Empirical Pilot Study. 2019 IEEE Jordan International Joint Conference on Electrical Engineering and Information Technology (JEEIT). :580–585.

In today's interconnected world, universities recognize the importance of protecting their information assets from internal and external threats. Being the possible insider threats to Information Security, employees are often coined as the weakest link. Both employees and organizations should be aware of this raising challenge. Understanding staff perception of compliance behaviour is critical for universities wanting to leverage their staff capabilities to mitigate Information Security risks. Therefore, this research seeks to get insights into staff perception based on factors adopted from several theories by using proposed constructs i.e. "perceived" practices/policies and "perceived" intention to comply. Drawing from the General Deterrence Theory, Protection Motivation Theory, Theory of Planned Behaviour and Information Reinforcement, within the context of Palestine universities, this paper integrates staff awareness of Information Security Policies (ISP) countermeasures as antecedents to ``perceived'' influencing factors (perceived sanctions, perceived rewards, perceived coping appraisal, and perceived information reinforcement). The empirical study is designed to follow a quantitative research approaches, use survey as a data collection method and questionnaires as the research instruments. Partial least squares structural equation modelling is used to inspect the reliability and validity of the measurement model and hypotheses testing for the structural model. The research covers ISP awareness among staff and seeks to assert that information security is the responsibility of all academic and administrative staff from all departments. Overall, our pilot study findings seem promising, and we found strong support for our theoretical model.

Irmak, E., Erkek, İ.  2018.  An overview of cyber-attack vectors on SCADA systems. 2018 6th International Symposium on Digital Forensic and Security (ISDFS). :1–5.

Most of the countries evaluate their energy networks in terms of national security and define as critical infrastructure. Monitoring and controlling of these systems are generally provided by Industrial Control Systems (ICSs) and/or Supervisory Control and Data Acquisition (SCADA) systems. Therefore, this study focuses on the cyber-attack vectors on SCADA systems to research the threats and risks targeting them. For this purpose, TCP/IP based protocols used in SCADA systems have been determined and analyzed at first. Then, the most common cyber-attacks are handled systematically considering hardware-side threats, software-side ones and the threats for communication infrastructures. Finally, some suggestions are given.

Isaacson, D. M..  2018.  The ODNI-OUSD(I) Xpress Challenge: An Experimental Application of Artificial Intelligence Techniques to National Security Decision Support. 2018 IEEE 8th Annual Computing and Communication Workshop and Conference (CCWC). :104-109.
Current methods for producing and disseminating analytic products contribute to the latency of relaying actionable information and analysis to the U.S. Intelligence Community's (IC's) principal customers, U.S. policymakers and warfighters. To circumvent these methods, which can often serve as a bottleneck, we report on the results of a public prize challenge that explored the potential for artificial intelligence techniques to generate useful analytic products. The challenge tasked solvers to develop algorithms capable of searching and processing nearly 15,000 unstructured text files into a 1-2 page analytic product without human intervention; these analytic products were subsequently evaluated and scored using established IC methodologies and criteria. Experimental results from this challenge demonstrate the promise for the ma-chine generation of analytic products to ensure that the IC warns and informs in a more timely fashion.
Isaakidis, Marios, Halpin, Harry, Danezis, George.  2016.  UnlimitID: Privacy-Preserving Federated Identity Management Using Algebraic MACs. Proceedings of the 2016 ACM on Workshop on Privacy in the Electronic Society. :139–142.

UnlimitID is a method for enhancing the privacy of commodity OAuth and applications such as OpenID Connect, using anonymous attribute-based credentials based on algebraic Message Authentication Codes (aMACs). OAuth is one of the most widely used protocols on the Web, but it exposes each of the requests of a user for data by each relying party (RP) to the identity provider (IdP). Our approach allows for the creation of multiple persistent and unlinkable pseudo-identities and requires no change in the deployed code of relying parties, only in identity providers and the client.

Isaeva, N. A..  2018.  Choice of Control Parameters of Complex System on the Basis of Estimates of the Risks. 2018 Eleventh International Conference "Management of Large-Scale System Development" (MLSD. :1-4.

The method of choice the control parameters of a complex system based on estimates of the risks is proposed. The procedure of calculating the estimates of risks intended for a choice of rational managing directors of influences by an allocation of the group of the operating factors for the set criteria factor is considered. The purpose of choice of control parameters of the complex system is the minimization of an estimate of the risk of the functioning of the system by mean of a solution of a problem of search of an extremum of the function of many variables. The example of a choice of the operating factors in the sphere of intangible assets is given.

Isah, H., Neagu, D., Trundle, P..  2015.  Bipartite network model for inferring hidden ties in crime data. 2015 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM). :994–1001.

Certain crimes are difficult to be committed by individuals but carefully organised by group of associates and affiliates loosely connected to each other with a single or small group of individuals coordinating the overall actions. A common starting point in understanding the structural organisation of criminal groups is to identify the criminals and their associates. Situations arise in many criminal datasets where there is no direct connection among the criminals. In this paper, we investigate ties and community structure in crime data in order to understand the operations of both traditional and cyber criminals, as well as to predict the existence of organised criminal networks. Our contributions are twofold: we propose a bipartite network model for inferring hidden ties between actors who initiated an illegal interaction and objects affected by the interaction, we then validate the method in two case studies on pharmaceutical crime and underground forum data using standard network algorithms for structural and community analysis. The vertex level metrics and community analysis results obtained indicate the significance of our work in understanding the operations and structure of organised criminal networks which were not immediately obvious in the data. Identifying these groups and mapping their relationship to one another is essential in making more effective disruption strategies in the future.

Isakov, M., Bu, L., Cheng, H., Kinsy, M. A..  2018.  Preventing Neural Network Model Exfiltration in Machine Learning Hardware Accelerators. 2018 Asian Hardware Oriented Security and Trust Symposium (AsianHOST). :62–67.

Machine learning (ML) models are often trained using private datasets that are very expensive to collect, or highly sensitive, using large amounts of computing power. The models are commonly exposed either through online APIs, or used in hardware devices deployed in the field or given to the end users. This provides an incentive for adversaries to steal these ML models as a proxy for gathering datasets. While API-based model exfiltration has been studied before, the theft and protection of machine learning models on hardware devices have not been explored as of now. In this work, we examine this important aspect of the design and deployment of ML models. We illustrate how an attacker may acquire either the model or the model architecture through memory probing, side-channels, or crafted input attacks, and propose (1) power-efficient obfuscation as an alternative to encryption, and (2) timing side-channel countermeasures.

Iscen, Ahmet, Furon, Teddy.  2016.  Group Testing for Identification with Privacy. Proceedings of the 4th ACM Workshop on Information Hiding and Multimedia Security. :51–56.

This paper describes an approach where group testing helps in enforcing security and privacy in identification. We detail a particular scheme based on embedding and group testing. We add a second layer of defense, group vectors, where each group vector represents a set of dataset vectors. Whereas the selected embedding poorly protects the data when used alone, the group testing approach makes it much harder to reconstruct the data when combined with the embedding. Even when curious server and user collude to disclose the secret parameters, they cannot accurately recover the data. Another byproduct of our approach is that it reduces the complexity of the search and the required storage space. We show the interest of our work in a benchmark biometrics dataset, where we verify our theoretical analysis with real data.

Ishak, Muhammad Yusry Bin, Ahmad, Samsiah Binti, Zulkifli, Zalikha.  2019.  Iot Based Bluetooth Smart Radar Door System Via Mobile Apps. 2019 1st International Conference on Artificial Intelligence and Data Sciences (AiDAS). :142—145.
{In the last few decades, Internet of things (IOT) is one of the key elements in industrial revolution 4.0 that used mart phones as one of the best technological advances' intelligent device. It allows us to have power over devices without people intervention, either remote or voice control. Therefore, the “Smart Radar Door “system uses a microcontroller and mobile Bluetooth module as an automation of smart door lock system. It is describing the improvement of a security system integrated with an Android mobile phone that uses Bluetooth as a wireless connection protocol and processing software as a tool in order to detect any object near to the door. The mob ile device is required a password as authentication method by using microcontroller to control lock and unlock door remotely. The Bluetooth protocol was chosen as a method of communication between microcontroller and mobile devices which integrated with many Android devices in secured protocol}.
Ishak, Z., Rajendran, N., Al-Sanjary, O. I., Razali, N. A. Mat.  2020.  Secure Biometric Lock System for Files and Applications: A Review. 2020 16th IEEE International Colloquium on Signal Processing Its Applications (CSPA). :23–28.

A biometric system is a developing innovation which is utilized in different fields like forensics and security system. Finger recognition is the innovation that confirms the personality of an individual which relies upon the way that everybody has unique fingerprints. Fingerprint biometric systems are smaller in size, simple to utilize and have low power. This proposed study focuses on fingerprint biometric systems and how such a system would be implemented. If implemented, this system would have multifactor authentication strategies and improvised features based on encryption algorithms. The scanner that will be used is Biometric Fingerprint Sensor that is connected to system which determines the authorization and access control rights. All user access information is gathered by the system where the administrators can retrieve and analyse the information. This system has function of being up to date with the data changes like displaying the name of the individual for controlling security of the system.

Ishaque, Mohammed, Hudec, Ladislav.  2019.  Feature extraction using Deep Learning for Intrusion Detection System. 2019 2nd International Conference on Computer Applications Information Security (ICCAIS). :1–5.

Deep Learning is an area of Machine Learning research, which can be used to manipulate large amount of information in an intelligent way by using the functionality of computational intelligence. A deep learning system is a fully trainable system beginning from raw input to the final output of recognized objects. Feature selection is an important aspect of deep learning which can be applied for dimensionality reduction or attribute reduction and making the information more explicit and usable. Deep learning can build various learning models which can abstract unknown information by selecting a subset of relevant features. This property of deep learning makes it useful in analysis of highly complex information one which is present in intrusive data or information flowing with in a web system or a network which needs to be analyzed to detect anomalies. Our approach combines the intelligent ability of Deep Learning to build a smart Intrusion detection system.

Ishiguro, Kenta, Kono, Kenji.  2018.  Hardening Hypervisors Against Vulnerabilities in Instruction Emulators. Proceedings of the 11th European Workshop on Systems Security. :7:1–7:6.

Vulnerabilities in hypervisors are crucial in multi-tenant clouds and attractive for attackers because a vulnerability in the hypervisor can undermine all the virtual machine (VM) security. This paper focuses on vulnerabilities in instruction emulators inside hypervisors. Vulnerabilities in instruction emulators are not rare; CVE-2017-2583, CVE-2016-9756, CVE-2015-0239, CVE-2014-3647, to name a few. For backward compatibility with legacy x86 CPUs, conventional hypervisors emulate arbitrary instructions at any time if requested. This design leads to a large attack surface, making it hard to get rid of vulnerabilities in the emulator. This paper proposes FWinst that narrows the attack surface against vulnerabilities in the emulator. The key insight behind FWinst is that the emulator should emulate only a small subset of instructions, depending on the underlying CPU micro-architecture and the hypervisor configuration. FWinst recognizes emulation contexts in which the instruction emulator is invoked, and identifies a legitimate subset of instructions that are allowed to be emulated in the current context. By filtering out illegitimate instructions, FWinst narrows the attack surface. In particular, FWinst is effective on recent x86 micro-architectures because the legitimate subset becomes very small. Our experimental results demonstrate FWinst prevents existing vulnerabilities in the emulator from being exploited on Westmere micro-architecture, and the runtime overhead is negligible.

Ishikawa, Tomohisa, Sakurai, Kouichi.  2017.  A Proposal of Event Study Methodology with Twitter Sentimental Analysis for Risk Management. Proceeding IMCOM '17 Proceedings of the 11th International Conference on Ubiquitous Information Management and Communication Article No. 14 .

Once organizations have the security incident and breaches, they have to pay tremendous costs. Although visible cost, such as the incident response cost, customer follow-up care, and legal cost are predictable and calculable, it is tough to evaluate and estimate the invisible damage, such as losing customer loyalty, reputation impact, and the damage of branding. This paper proposes a new method, called "Event Study Methodology with Twitter Sentimental Analysis" to evaluate the invisible cost. This method helps to assess the impact of the security breach and the impact on corporate valuation.

 

Ishio, T., Sakaguchi, Y., Ito, K., Inoue, K..  2017.  Source File Set Search for Clone-and-Own Reuse Analysis. 2017 IEEE/ACM 14th International Conference on Mining Software Repositories (MSR). :257–268.
Clone-and-own approach is a natural way of source code reuse for software developers. To assess how known bugs and security vulnerabilities of a cloned component affect an application, developers and security analysts need to identify an original version of the component and understand how the cloned component is different from the original one. Although developers may record the original version information in a version control system and/or directory names, such information is often either unavailable or incomplete. In this research, we propose a code search method that takes as input a set of source files and extracts all the components including similar files from a software ecosystem (i.e., a collection of existing versions of software packages). Our method employs an efficient file similarity computation using b-bit minwise hashing technique. We use an aggregated file similarity for ranking components. To evaluate the effectiveness of this tool, we analyzed 75 cloned components in Firefox and Android source code. The tool took about two hours to report the original components from 10 million files in Debian GNU/Linux packages. Recall of the top-five components in the extracted lists is 0.907, while recall of a baseline using SHA-1 file hash is 0.773, according to the ground truth recorded in the source code repositories.
Ishtiaq, Asra, Islam, Muhammad Arshad, Azhar Iqbal, Muhammad, Aleem, Muhammad, Ahmed, Usman.  2019.  Graph Centrality Based Spam SMS Detection. 2019 16th International Bhurban Conference on Applied Sciences and Technology (IBCAST). :629–633.

Short messages usage has been tremendously increased such as SMS, tweets and status updates. Due to its popularity and ease of use, many companies use it for advertisement purpose. Hackers also use SMS to defraud users and steal personal information. In this paper, the use of Graphs centrality metrics is proposed for spam SMS detection. The graph centrality measures: degree, closeness, and eccentricity are used for classification of SMS. Graphs for each class are created using labeled SMS and then unlabeled SMS is classified using the centrality scores of the token available in the unclassified SMS. Our results show that highest precision and recall is achieved by using degree centrality. Degree centrality achieved the highest precision i.e. 0.81 and recall i.e., 0.76 for spam messages.

Iskhakov, A., Jharko, E..  2020.  Approach to Security Provision of Machine Vision for Unmanned Vehicles of “Smart City”. 2020 International Conference on Industrial Engineering, Applications and Manufacturing (ICIEAM). :1—5.

By analogy to nature, sight is the main integral component of robotic complexes, including unmanned vehicles. In this connection, one of the urgent tasks in the modern development of unmanned vehicles is the solution to the problem of providing security for new advanced systems, algorithms, methods, and principles of space navigation of robots. In the paper, we present an approach to the protection of machine vision systems based on technologies of deep learning. At the heart of the approach lies the “Feature Squeezing” method that works on the phase of model operation. It allows us to detect “adversarial” examples. Considering the urgency and importance of the target process, the features of unmanned vehicle hardware platforms and also the necessity of execution of tasks on detecting of the objects in real-time mode, it was offered to carry out an additional simple computational procedure of localization and classification of required objects in case of crossing a defined in advance threshold of “adversarial” object testing.