Larus Technologies (Canada)

ABSTRACT Sources of light characterized by a well‐defined number of photons are widely used in quantum experiments and technologies due to their peculiar properties, which include sub‐Poissonian statistics, spatial, temporal, and frequency correlations, as well as maximal and high‐dimensional entanglement. Searching for the best experimental approaches that allow us to visualize the quantum state of two (or more) photons has been a long‐standing fundamental question, which has been mainly addressed via mode projection techniques. The rapid development of single‐photon‐sensitive cameras, however, has opened the pathway to conceptually simple, yet faster and more efficient, measurement techniques. This review explores the latest advancements in measuring the spatial structure of the quantum state of light using 3D imaging techniques. An overview of the most used single‐photon camera technologies is given, highlighting their differences and respective advantages. Besides the fundamental interest in reconstructing experimentally one of the most mysterious concepts of microscopic physics, this review illustrates how the techniques developed in this direction can lead to new ideas in the fields of imaging and sensing, for instance, superresolution measurements and phase‐enhanced and sub‐shot noise imaging.

Border and security agencies often rely on the Automatic Identification System (AIS) to gather intelligence and information on the current situation in their waters. However, they cannot always rely on the ship's AIS transceiver to operate properly. Using Convolutional Neural Networks (CNNs) for the detection of vessels in images and Field Programmable Gate Arrays (FPGA) platforms for deployment, it is possible to deploy a solution that is both faster than Graphical Processing Units (GPUs) and much more power efficient. This solution provides authorities with more tools to secure their ports. In this article, we propose an object detection solution capable of identifying individual ships from images as well as its Register Transfer Level (RTL) design that offers a speedup on inference time by a factor of 4 as well as consuming almost 50 times less power than the traditional GPU solution.

Maritime trade accounts for approximately 90% of global trade, and most global supply chains incorporate some form of maritime travel (Cheraghchi et al., 2017). Optimising operations at commercial maritime ports is therefore of significant importance worldwide, and impacts global trade. While damage to vessels and cargo has been studied extensively, as has optimising portside operational efficiency, investigations of damage to shipping containers themselves and the resultant disruption of port-side efficiency remains unstudied. The application of machine learning (ML) techniques to uncover causes of shipping container damage allows for more efficient handling of the service-disruptions they cause, as well as insights into the veracity of the current wisdom held by domain experts in the industry. Further, the application of ML methodologies for dynamic algorithm selection (per level 4 of the JDL/DFIG data fusion model) allows for significant improvements to the overall performance of the aforementioned ML methodologies.

A recent trend in Machine Learning is to augment the transparency of traditional classification models using Granular Computing techniques. This approach has been found particularly useful in the neural networks field since most successful neural systems often require complex structures to behave like universal approximators. However, there is a widely-held view stating that, to build an interpretable classifier, one might have to sacrifice some prediction accuracy. We want to challenge this belief by exploring the performance of a recently introduced granular classifier termed Fuzzy-Rough Cognitive Networks against low-level (i.e., traditional) neural networks. The simulation results have shown that this neural system can attain quite competitive prediction rates while featuring a shallow, granular architecture. As a bigger picture, this study paves the way for conducting a more thorough evaluation of granular versus low-level neural classifiers in the near future.

With the rise of more resourceful unmanned aerial vehicles (UAVs), their inclusion into robotic sensor networks (RSNs) is inevitable. The highly mobile nature of UAVs allows greater monitoring capabilities, making them most suitable for RSNs. Compared to traditional nodes in RSNs, UAVs suffer even more from communication disruptions and energy depletion, must often rapidly determine actions for themselves, and consequently require more autonomy. Prior work has been done in wireless sensor network (WSN)/aerial sensor network (ASN) coordination in a few applications such as protecting critical infrastructure, restoring communication between nodes, and healing networks, while other work has been accomplished on using the UAV network for augmenting the monitoring capabilities of WSNs. We introduce a novel methodology to integrate UAVs into RSNs for monitoring purposes by formulating the problem in the context of a risk management framework (RMF). This methodology allows a more precise risk feature classification and a more efficient task allocation for the ground network by utilizing the monitoring capabilities of the UAVs to informatively warn the RSN of any incoming events. We also present a fictitious but credible maritime smuggling scenario near the Port of Barcelona based on expert knowledge, and apply the methodology to detect and mitigate maritime smuggling. The network's behaviour is traced throughout the scenario and is repeated with civilian ships to assure that they are not flagged as smugglers. The applied methodology results in a successful classification and mitigation of the smuggling activity.

Cet article analyse les relations entre docteurs et spirituels dans l’Espagne moderne. Au xvie siècle, les docteurs ont fini par imposer leur point de vue. Trois dates marquent les étapes de cette évolution : la mort de Cisneros en 1517, l’édit de l’Inquisition contre les alumbrados de Tolède en 1525, et l’index de l’inquisiteur général Valdés en 1559.

Individual units in a wireless sensor network (WSN) are exposed to multiple risks, either during or after their deployment. The identification of the risk sources and their watchful monitoring in dynamic, unpredictable environments is pivotal to ensure a smooth, long-term functioning of the WSN. We introduce an evolving risk management framework for WSNs that captures multiple risk features and provides both a visual depiction of the corporate network threats at any time and a numerical assessment of any sensor's overall risk. The visualization module is embodied through an evolving clustering architecture which heavily relies on shadowed sets. The risk assessment module embraces fuzzy and shadowed evaluations of the risk sources and incorporates a simple adaptive learning process that weights the risk sources proportionally to their observed impact on failed sensors. A distinctive trait of the proposed framework is its highly automated yet still human-centric nature. Experiments utilizing different sensor models and deployment scenarios confirm the feasibility of the risk management platform under consideration.

This paper introduces a novel architecture that is targeted for digital core testing and built-in-self test (BIST) algorithm development. This reconfigurable architecture is validated by an application that implements the novel idea of verifying algorithms for testing digital circuits by using runtime reconfigurable techniques in order to minimize the circuit area, as well as the test generation and application time. The idea revolves around the dynamic partial reconfiguration of circuits under test in order to inject stuck-at faults at different locations of the circuit and uncover both detectable and undetectable faults. Four testing strategies are presented, and two are experimentally compared, namely, the sequential compile-time reconfiguration and runtime reconfiguration strategies.

This paper presents an overview of infrastructure-based localization schemes for indoor wireless sensor networks. Five areas were identified that indiscriminately delineate one sensor network from another, as well as eight classification properties for the different localization schemes. Finally, we present a novel localization methodology that does not require the sensor node to perform any on-board calculations in order to infer its location, and its associated metrics.