Abstract: We present a first-of-a-kind end-to-end framework for running privacy risk assessments of AI models that enables assessing models from multiple ML frameworks, using a variety of low-level privacy attacks and metrics. The tool automatically selects which attacks and metrics to run based on answers to questions, runs the attacks, summarizes and visualizes the results in an easy-to-consume manner.

• acm.org
  Visit

• No downloads available

Abstract: We are currently seeing an increase in the use of voice assistants which are used for various purposes. These assistants have a wide range of inbuilt functionalities with the possibility of installing third-party applications. In this work, we will focus on analyzing and identifying vulnerabilities that are introduced by these third-party applications. In particular, we will build third-party applications (called Skills) for Alexa, the voice assistant developed by Amazon. We will analyze existing exploits, identify accessible data and propose an adversarial framework that deceives users into disclosing private information. For this purpose, we developed four different malicious Skills that harvest different pieces of private information from users. We perform a usability analysis on the Skills and feasibility analysis on the publishing pipeline for one of the Skills.

• cyberkit4sme.eu
  Visit
• ac.uk
  Visit

• No downloads available

Abstract: Small and medium enterprises (SMEs) make up a significant part of European economies. They are often described as poorly place to deal with cyber risks though because of resource constraints or commercial interests. Providing appropriate tooling would facilitate a greater appreciation of the risks and provide mitigation strategies. In a series of workshops demonstrating visualization tools for cybersecurity, constructs from healthcare models such as awareness, self-efficacy, and a willingness to engage were investigated to throw light on the likelihood that the technologies would be adopted. Although most constructs were validated, it turns out that self-efficacy could more appropriately be interpreted as a desire to understand a broader company narrative rather than empowering any individual to identify and manage cyber risk. As part of an ongoing examination of technology acceptance, this work provides further evidence that technology must be contextualized to make sense for the individual as part of the SME rather than as individual employee.

• springer.com
  Visit

• Pickering_etal_ItsNotMyProblem_WorkingCopy_02022021
  Download

Abstract: Forest roads in Romania are unique natural wildlife sites used for recreation by countless tourists. In order to protect and maintain these roads, we propose RovisLab AMTU (Autonomous Mobile Test Unit), which is a robotic system designed to autonomously navigate off-road terrain and inspect if any deforestation or damage occurred along tracked route. AMTU’s core component is its embedded vision module, optimized for real-time environment perception. For achieving a high computation speed, we use a learning system to train a multi-task Deep Neural Network (DNN) for scene and instance segmentation of objects, while the keypoints required for simultaneous localization and mapping are calculated using a handcrafted FAST feature detector and the Lucas-Kanade tracking algorithm. Both the DNN and the handcrafted backbone are run in parallel on the GPU of an NVIDIA AGX Xavier board. We show experimental results on the test track of our research facility.

• arxiv.org
  Visit

• No downloads available

Abstract: Small and medium-sized enterprises (SMEs) rarely conduct a thorough cyber-risk assessment and they may face various internal issues when attempting to set up cyber-risk strategies. In this work, we apply a user journey approach to model human behaviour and visually map SMEs’ practices and threats, along with a visualisation of the socio-technical actor network, targeted specifically at the risks highlighted in the user journey. By using a combination of cybersecurity-related visualisations, our goals are: i) to raise awareness about cybersecurity, and ii) to improve communication among IT personnel, security experts, and non-technical personnel. To achieve these goals, we combine two modelling languages: Customer Journey Modelling Language (CJML) is a visual language for modelling and visualisation of work processes in terms of user journeys. System Security Modeller (SSM) is an asset-based risk-analysis tool for socio-technical systems. By demonstrating the languages’ supplementary nature through a threat scenario and considering related theories, we believe that there is a sound basis to warrant further validation of CJML and SSM together to raise awareness and handle cyber threats in SMEs.

• scitepress.org
  Visit
• boletsis.net
  Visit

• No downloads available

Abstract: Self-driving cars and autonomous vehicles are revolutionizing the automotive sector, shaping the future of mobility altogether. Although the integration of novel technologies such as Artificial Intelligence (AI) and Cloud/Edge computing provides golden opportunities to improve autonomous driving applications, there is the need to modernize accordingly the whole prototyping and deployment cycle of AI components. This paper proposes a novel framework for developing so-called AI Inference Engines for autonomous driving applications based on deep learning modules, where training tasks are deployed elastically over both Cloud and Edge resources, with the purpose of reducing the required network bandwidth, as well as mitigating privacy issues. Based on our proposed data driven V-Model, we introduce a simple yet elegant solution for the AI components development cycle, where prototyping takes place in the cloud according to the Software-in-the-Loop (SiL) paradigm, while deployment and evaluation on the target ECUs (Electronic Control Units) is performed as Hardware-in-the-Loop (HiL) testing. The effectiveness of the proposed framework is demonstrated using two real-world use-cases of AI inference engines for autonomous vehicles, that is environment perception and most probable path prediction.

• arxiv.org
  Visit
• mdpi.com
  Visit

• No downloads available