General Scenario Research Articles

A Lightweight Object Detector Based on Spatial-Coordinate Self-Attention for UAV Aerial Images

Object detection is one of the most widespread applications for numerous Unmanned Aerial Vehicle (UAV) tasks. Due to the shooting angle and flying height of the UAV, compared with general scenarios, small objects account for a large proportion of aerial images, and common object detectors are not extremely effective in aerial images. Moreover, since the computing resources of UAV platforms are generally limited, the deployment of common detectors with a large number of parameters on UAV platforms is difficult. This paper proposes a lightweight object detector YOLO-UAVlite for aerial images. Firstly, the spatial attention module and coordinate attention module are modified and combined to form a novel Spatial-Coordinate Self-Attention (SCSA) module, which integrates spatial, location, and channel information to enhance object representation. On this basis, we construct a lightweight backbone, named SCSAshufflenet, which combines the Enhanced ShuffleNet (ES) network with the proposed SCSA module to improve feature extraction and reduce model size. Secondly, we propose an improved feature pyramid model, namely Slim-BiFPN, where we construct new lightweight convolutional blocks to reduce the information loss during the feature map fusion process while reducing the model weights. Finally, the localization loss function is modified to increase the bounding box regression rate while improving the localization accuracy. Extensive experiments conducted on the VisDrone-DET2021 dataset indicate that, compared with the YOLOv5-N baseline, the proposed YOLO-UAVlite reduces the number of parameters by 25.8% and achieves gains of 10.9% in mAP0.50. Compared with other lightweight detectors, both the mAP and the number of parameters are improved.

Medical knowledge-based network for Patient-oriented Visual Question Answering

Visual Question Answering (VQA) systems have achieved great success in general scenarios. In medical domain, VQA systems are still in their infancy as the datasets are limited by scale and application scenarios. Current medical VQA datasets are designed to conduct basic analyses of medical imaging such as modalities, planes, organ systems, abnormalities, etc., aiming to provide constructive medical suggestions for doctors, containing a large number of professional terms with limited help for patients. In this paper, we introduce a new Patient-oriented Visual Question Answering (P-VQA) dataset, which builds a VQA system for patients by covering an entire treatment process including medical consultation, imaging diagnosis, clinical diagnosis, treatment advice, review, etc. P-VQA covers 20 common diseases with 2,169 medical images, 24,800 question-answering pairs, and a medical knowledge graph containing 419 entities. In terms of methodology, we propose a Medical Knowledge-based VQA Network (MKBN) to answer questions according to the images and a medical knowledge graph in our P-VQA. MKBN learns two cluster embeddings (disease-related and relation-related embeddings) according to structural characteristics of the medical knowledge graph and learns three different interactive features (image-question, image-disease, and question-relation) according to characteristics of diagnosis. For comparisons, we evaluate several state-of-the-art baselines on the P-VQA dataset as benchmarks. Experimental results on P-VQA demonstrate that MKBN achieves the state-of-the-art performance compared with baseline methods. The dataset is available at https://github.com/cs-jerhuang/P-VQA.

Data-Driven Intelligent Manipulation of Particles in Microfluidics.

Automated manipulation of small particles using external (e.g., magnetic, electric and acoustic) fields has been an emerging technique widely used in different areas. The manipulation typically necessitates a reduced-order physical model characterizing the field-driven motion of particles in a complex environment. Such models are available only for highly idealized settings but are absent for a general scenario of particle manipulation typically involving complex nonlinear processes, which has limited its application. In this work, the authorspresent a data-driven architecture for controlling particles in microfluidics based on hydrodynamic manipulation. The architecture replaces the difficult-to-derive model by a generally trainable artificial neural network to describe the kinematics of particles, and subsequently identifies the optimal operations to manipulate particles. The authorssuccessfully demonstrate a diverse set of particle manipulations in a numerically emulated microfluidic chamber, including targeted assembly of particles and subsequent navigation of the assembled cluster, simultaneous path planning for multiple particles, and steering one particle through obstacles. Theapproach achieves both spatial and temporal controllability of high precision for these settings. This achievement revolutionizes automated particle manipulation, showing the potential of data-driven approaches and machine learning in improving microfluidic technologies for enhanced flexibility andintelligence.

Quantum sets of the multicolored-graph approach to contextuality

The CHSH inequalities are the most famous examples of Bell inequalities. Cabello, Severini, and Winter (CSW) came up with a graph approach to noncontextuality inequalities, which connects some graph-theoretic concepts to quantum and classical correlations. For example, the theta body of the exclusivity graph can be associated with the set of correlations achieved by quantum theory. Following the CSW approach, one may think that the theta body of the CHSH graph, is equal to the quantum set of the CHSH Bell inequality, but is this really true? All assumptions about the CHSH inequalities come from Bell scenarios, while CSW approach only demands the exclusivity structure of a non-contextuality (NC) scenario. To deal with the extra structure related to the presence of different players in a Bell scenario like CHSH, the colored-graph approach was introduced. Does it make any difference to think about CHSH as a Bell scenario or a more general NC scenario? The Bell CHSH inequality is represented by a bicolored graph and the NC CHSH inequality by a simple graph, which is the shadow of the colored graph. In general, we have that the theta body of the colored graph is a subset of the theta body of its shadow graph in the same way that the Lov\`asz number, which corresponds to the quantum bound, of the simple graph is greater than or equal to the Lov\`asz number of the colored graph. In the case of the CHSH inequality, we have that the Lov\`asz numbers are equal. Does this accident also hold for the corresponding quantum sets? Is it true that the theta bodies are equal, which would mean that every correlation reached by quantum theory applied to the CHSH NC scenario could also be obtained at the in principle more restrictive CHSH Bell scenario? In this paper our answer to such a question is negativ. This implies that there are quantum correlations which can not be obtained under Bell restrictions.

P2SIM: Privacy-Preserving and Source-Reliable Incentive Mechanism for Mobile Crowdsensing

In mobile crowdsensing (MCS), providing appropriate rewards is a common and efficient way to motivate participants to participate in sensing tasks. However, the privacy of task participants is not protected well in most quality-aware incentive schemes. Moreover, these schemes are designed for general MCS application scenarios where data are collected by internal sensors embedded in participants’ smartphones, and not suitable for scenarios where additional sensors (ASs) except internal sensors are to collect data (e.g., household medical devices). In scenarios with ASs, malicious participants can fabricate sensing data instead of collecting data from ASs, i.e., the source reliability of sensing data cannot be ensured. To address these issues, we propose P2SIM, a privacy-preserving and the source-reliable incentive mechanism scheme for MCS with ASs. We combine redactable signature with private hash function to achieve the source reliability verification of sensing data without revealing the privacy of participants. Moreover, rewards are divided into two parts: 1) fixed rewards and 2) floating rewards, to enhance the flexibility of rewards distribution. Both formal theoretical analysis and extensive experimental evaluations on a real data set show that the proposed P2SIM is secure and efficient.

Predicting young drivers’ safe behaviour of stopping in the dilemma zone

Young drivers are overrepresented in road crashes, half of which occur at intersections in urban areas. One area of focus for crashes at intersections is the dilemma zone, when a driver approaches a signalised intersection at the onset of the yellow light and is unsure whether to continue through or to stop. A challenge in this area of research has been the lack of focus on safe, rather than risky, driving behaviours of young drivers at intersections and a paucity of research on psychosocial factors that motivate driving decisions at intersections. This study aimed to explain young drivers’ intentions and predict their self-reported behaviour of stopping in the dilemma zone (the safe behaviour), using psychosocial factors which influence behaviour in young drivers. Participants were N = 180 young drivers from Queensland. An initial online survey presented three unique scenarios to participants: a general dilemma zone situation, the dilemma zone whilst running late and the dilemma zone where the vehicle in front drives through the yellow light. Participants responded to items measuring Theory of Planned Behaviour (TPB) constructs extended with prototype perceptions and willingness constructs from the Prototype Willingness Model (PWM). A follow-up online survey measured participants’ self-reported driving behaviours in the previous three weeks, in each scenario. Hierarchical multiple regressions found that the TPB and PWM constructs significantly explained between 59.1 % and 72.7 % of the variance in intentions to stop. This suggested the utility of extending the TPB with a social-reaction decision making model such as the PWM. The follow-up survey (N = 74), found that intentions, but not PBC or willingness, significantly predicted 29 % of the variance in young drivers’ self-reported stopping behaviour for the general scenario and the running late scenario and 10.8 % for the vehicle in front scenario. The findings of this study suggest the value of researching young drivers’ safe behaviours and provide information on psychosocial factors, such as prototype perceptions, that road safety interventions can target in order to reduce road crashes. Additionally, these findings provide evidence for the use of the PWM to extend the TPB, and in investigating young drivers’ decisions in the dilemma zone.

Improving the realism of neutral ecological models by incorporating transient dynamics with temporal changes in community size

Neutral models in ecology assume that all species are demographically equivalent, such that their abundances differ ultimately because of demographic stochasticity rather than selection. In spite of their simplicity, neutral models have been found to accurately reproduce static patterns of biodiversity for diverse communities. However, the same neutral models have been found to exhibit species abundance dynamics that are far too slow compared to reality, resulting in poor fits to temporally dynamic patterns of biodiversity. Here, we show that one of the root causes of these slow dynamics is the additional assumption that a community has reached an equilibrium with a fixed community size, with species that have a net growth rate close to zero. We removed this additional assumption by constructing and analyzing a neutral model with an expected community size that can change over time and is not necessarily at equilibrium, which thus allows the historical formation of a community to be represented explicitly. Our analysis demonstrated that for the general scenario where a small community rapidly grows in size to a carrying capacity, representing recovery from ecological disturbance or assembly of a new community, the model produced much larger changes in species abundances and much shorter species ages than a neutral model at an equilibrium with fixed community size. In addition, the species abundance distribution was biphasic with a subset of abundant species arising from a founder effect. We confirmed these new results in applications of the new model to the specific scenario of recovery of the Amazon tree community after the end-Cretaceous bolide impact, which involved periods of increasing and decreasing community size. We conclude that incorporating transient dynamics in neutral models improves realism by allowing explicit consideration of how a community is formed over realistic time-scales.

Statistical analysis of PV penetration impact on residential distribution grids

This paper presents a comprehensive approach to the probabilistic analysis of residential distribution grid injected by distributed PV sources. The approach is data-driven and is able to deal with the general scenario that includes the uncertainty of correlated PV sources and of statistically independent consumer loads. The novel method adopts a Gaussian-Mixture-Model for correctly representing PV sources correlation and a fresh probabilistic analysis technique employing Multi-Expansion polynomial chaos. Numerical experiments carried out on the Non-Synthetic European low voltage test system, highlight the importance of the comprehensive modeling strategy in order to realistically quantify uncertainty impact on the grid.

Virtual Generator-Storage Pairing for Robust Multistage Decisions in Power Systems

In this article, we present a robust optimization-based framework for day-ahead reliability assessment (RA) and real-time dispatch to ensure power system reliability under multistage renewable uncertainty. Our framework jointly considers fossil-fuel generators and energy storage in the system. To obtain near-optimal performance in grid reliability, we develop robust multistage decisions based on the concept of “safe-dispatch sets” in the literature. Although such safe-dispatch sets directly answer the RA and real-time dispatch problem, their computation often incurs high computational complexity. To develop low-complexity algorithms, we adopt a divide-and-conquer paradigm. First, we derive the conditions to describe the safe-dispatch sets for a pair of energy storage and generator. The results from this simple building block provide us with useful insights on how the grid reliability is related to the parameters, e.g., the generators’ ramp speeds and the storage capacity. To tackle the general multibus scenario, the uncertainty is split among virtual generator-storage pairs (VGSPs), where we can utilize the results from the one-bus building block. Numerical studies on a standard IEEE 30-bus test case illustrate that our proposed solution requires much lower storage capacity to ensure system reliability compared to state-of-the-art approaches, without renewable curtailment.

MTLP-JR: Multi-task learning-based prediction for joint ranking in neural architecture search

At present, great attentions have been paid to multi-objective neural architecture search (NAS) and resource-aware NAS for their comprehensive consideration of the overall evaluation of architectures, including inference latency, precision, and model scale. However NAS also exacerbates the ever-increasing cost (engineering, time complexity, computation resource). Aiming to alleviate this, the reproducible NAS research releases the benchmark, which includes the metrics (e.g. Accuracy, Latency, and Parameters) of representative models from the typical search space on specific tasks. Motivated by the multi-objective NAS, resource-aware NAS, and reproducible NAS, this paper dedicates to binary-relation prediction (Latency, Accuracy), which is a more reasonable and effective way to satisfy the general NAS scenarios with less cost. We conduct a reproducible NAS study on the MobileNet-based search space and release the dataset. Further, we first propose the modeling of common features among prediction tasks (Latency, Accuracy, Parameters, and FLOPs), which will facilitate the prediction of individual tasks, and creatively formulate the architecture ranking prediction with a multi-task learning framework. Eventually, the proposed multi-task learning based binary-relation prediction model reaches the performance of 94.3% on Latency and 85.02% on Top1 Accuracy even with only 100 training points, which outperforms the single-task learning based model.

Detecting quantum capacities of continuous-variable quantum channels

Quantum communication channels and quantum memories are the fundamental building blocks of large-scale quantum communication networks. Estimating their capacity to transmit and store quantum information is crucial in order to assess the performance of quantum communication systems and to detect useful communication paths among the nodes of future quantum networks. However, the estimation of quantum capacities is a challenging task for continuous-variable systems, such as the radiation field, for which a complete characterization via quantum tomography is practically unfeasible. Here we introduce a method for detecting the quantum capacity of continuous-variable communication channels and memories without performing a full process tomography. Our method works in the general scenario where the devices are used a finite number of times, can exhibit correlations across multiple uses, and can change dynamically under the control of a malicious adversary. The method is experimentally friendly and can be implemented using only finitely squeezed states and homodyne measurements.

A multi-scale feature extraction fusion model for human activity recognition

Human Activity Recognition (HAR) is an important research area in human–computer interaction and pervasive computing. In recent years, many deep learning (DL) methods have been widely used for HAR, and due to their powerful automatic feature extraction capabilities, they achieve better recognition performance than traditional methods and are applicable to more general scenarios. However, the problem is that DL methods increase the computational cost of the system and take up more system resources while achieving higher recognition accuracy, which is more challenging for its operation in small memory terminal devices such as smartphones. So, we need to reduce the model size as much as possible while taking into account the recognition accuracy. To address this problem, we propose a multi-scale feature extraction fusion model combining Convolutional Neural Network (CNN) and Gated Recurrent Unit (GRU). The model uses different convolutional kernel sizes combined with GRU to accomplish the automatic extraction of different local features and long-term dependencies of the original data to obtain a richer feature representation. In addition, the proposed model uses separable convolution instead of classical convolution to meet the requirement of reducing model parameters while improving recognition accuracy. The accuracy of the proposed model is 97.18%, 96.71%, and 96.28% on the WISDM, UCI-HAR, and PAMAP2 datasets respectively. The experimental results show that the proposed model not only obtains higher recognition accuracy but also costs lower computational resources compared with other methods.

The perspective of a smart city by endorsing the nexus Bermuda triangle with the risk assessment of polluted water reuse in integrated water and food security management: the case of Semnan, Iran

AbstractLocated in a semi-arid and arid zone, Iran is suffering from growing challenges of water scarcity. In the paradigm of a circular economy, the reuse of treated wastewater in agriculture is currently regarded as a possible solution for alleviating the issues of water scarcity and pollution. Accordingly, this research aims to assess the use of polluted water in the integrated management of water resources in Semnan. The research used the Water Evaluation and Planning System (WEAP) software package to model and analyze the water sector. Also, the Bayesian network method was used to assess the risk of using polluted water and its effects on humans and plants. The research explored two general scenarios for the study site of Semnan. The first scenario assumes the increase in population, crops (food), and industries, and the second has the same assumptions plus an increase in agricultural efficiency (food production). Based on the results, the agricultural, urban, and industrial water demands are 37, 0.06, and 0.01 million m3 in the base year, respectively. The water demand in the next years will be higher due to population growth. Finally, it is safer to use the wastewater of both treatment plants of the region (Mehdishahr and Semnan) in the industry than in other sectors. Additionally, the wastewater of the Mehdishahr Sewage Treatment Plant is more reliable than that of the Semnan Sewage Treatment Plant.

Modeling and simulation of lane-changing and collision avoiding autonomous vehicles on superhighways

Collision avoidance during vehicle lane-changing was studied using modeling simulation verification method to solve the problem of collision of vehicles on superhighways. The lateral acceleration model and the collision avoidance minimum safety spacing model were established based on the theoretical analysis of the quintic polynomial lane-changing trajectory model. Additionally, the simulated minimum safety spacing of collision avoidance was calculated using vehicle speed and lane-changing time. The simulation scene was divided into general and emergency lane-changing scenarios under the condition of limiting the maximum lateral acceleration, and the simulation scenarios were subdivided into seven simulation scenarios according to different lane-changing times and road widths for simulation verification. The results demonstrated that a longer lane-changing time was selected in the general lane-changing scenario, and the lateral acceleration and swing angle speed of the vehicle decrease with the increase of the lane-changing time. This ensured collision avoidance, stable vehicle driving, and comfort and safety of passengers. The minimum safety spacing and lane-changing time were short in the emergency lane-changing scenario, and the vehicle exhibited a mild skidding phenomenon during the lane-changing. However, rollover did not occur, which ensured collision avoidance. The distances between the vehicles after the lane avoidance simulation were 4.87, 5.06, 6.89, 6.49, 7.76, 4.21, and 5.52 m, respectively. It demonstrated that the seven lane-changing and avoidance simulation models involved in this paper had good effect on vehicle collision avoidance. The effectiveness and accuracy of the steering collision avoidance of autonomous vehicles on superhighways are verified using the simulation platform, which can effectively improve the safety of superhighway driving.

An Anisotropic Elastic Solid With an Elliptical Inhomogeneity Under a Non-Uniform In-Plane And Anti-Plane Remote Loading

Summary We consider the case of an anisotropic elastic elliptical inhomogeneity embedded in an infinite anisotropic elastic matrix subjected to a non-uniform remote loading described by remote stresses and strains which are linear functions of the two in-plane coordinates. The internal stresses and strains within the elliptical inhomogeneity are found to be linear functions of the two in-plane coordinates. In addition, we obtain explicit real-form solutions describing the elastic field inside the inhomogeneity as well as hoop stress vectors and hoop stresses on the matrix side and on the inhomogeneity side. We also obtain the corresponding solutions for the two limiting cases in which the elliptical inhomogeneity takes the form of a hole or a rigid inhomogeneity. The solution method presented here can be extended to accommodate the more general scenario in which the remote applied stresses and strains are arbitrary-order polynomials of the two in-plane coordinates.

Universal construction of genuinely entangled subspaces of any size

We put forward a simple construction of genuinely entangled subspaces – subspaces supporting only genuinely multipartite entangled states – of any permissible dimensionality for any number of parties and local dimensions. The method uses nonorthogonal product bases, which are built from totally nonsingular matrices with a certain structure. We give an explicit basis for the constructed subspaces. An immediate consequence of our result is the possibility of constructing in the general multiparty scenario genuinely multiparty entangled mixed states with ranks up to the maximal dimension of a genuinely entangled subspace.

Blazars at Very High Energies: Emission Modelling

Blazars are very broadband cosmic sources with spectra spanning over twenty orders of magnitude in frequency, down to the 100 MHz regime in the radio range, up to VHE at several tens of TeV. The modelling of their spectral energy distribution at high energies currently considers two main classes of models, leptonic and lepto-hadronic, which both succeed fairly well in describing the observed spectra for the two populations of blazars, namely BL Lac objects (BL Lacs) and flat spectrum radio quasars (FSRQs). However they are both confronted with difficulties, in particular to reproduce flaring phenomena monitored with a good multi-spectral and temporal coverage, or to reproduce extreme sources which challenge the basic descriptions. Such a situation has led to a diversity of specific scenarios, the positioning of which in relation to the general context of the sources is generally not clearly fixed. The identification of the dominant particle acceleration mechanism at work and a better understanding of the location of the TeV emitting zone would make it possible to break the degeneracies between models. Multi-wavelength and multi-messenger studies should also help in this regard, with the perspective to elaborate a general reference scenario of blazars and AGNs.

Classically Replaceable Operations

Quantum information science provides powerful technologies beyond the scope of classical physics. In practice, accurate control of quantum operations is a challenging task with current quantum devices. The implementation of high fidelity and multi-qubit quantum operations consumes massive resources and requires complicated hardware design to fight against noise. An approach to alleviating this problem is to replace quantum operations with classical processing. Despite the common practice of this approach, rigorous criteria to determine whether a given quantum operation is replaceable classically are still missing. In this work, we define the classically replaceable operations in four general scenarios. In each scenario, we provide their necessary and sufficient criteria and point out the corresponding classical processing. For a practically favorable case of unitary classically replaceable operations, we show that the replaced classical processing is deterministic. Beyond that, we regard the irreplaceability of quantum operations by classical processing as a quantum resource and relate it to the performance of a channel in a non-local game, as manifested in a robustness measure.

Taming the long distance effects in the Ds+→π+ℓ−ℓ+ decay

The semileptonic decay ${D}_{s}^{+}\ensuremath{\rightarrow}{\ensuremath{\pi}}^{+}{\ensuremath{\ell}}^{\ensuremath{-}}{\ensuremath{\ell}}^{+}$, where $\ensuremath{\ell}=e$, $\ensuremath{\mu}$, is used as a reference channel when looking for suppressed or forbidden processes of the standard model of the form ${D}_{s}\ensuremath{\rightarrow}P{\ensuremath{\ell}}^{\ifmmode\pm\else\textpm\fi{}}{\ensuremath{\ell}}^{\ensuremath{'}+}$. The process is dominated by hadronic resonances from the $\ensuremath{\eta}$ to the $\ensuremath{\phi}$ meson region, usually excluded due to the large uncertainties associated. In this work, we explore the role of the parameters involved in the resonance region. We focus in the relative strong phase between the $\ensuremath{\rho}$ and the $\ensuremath{\phi}$ meson (${\ensuremath{\delta}}_{\ensuremath{\rho}\ensuremath{\phi}}$). We argue that this parameter can be bound from LHCb data of the dimuon invariant mass in the ${D}_{s}^{+}\ensuremath{\rightarrow}{\ensuremath{\pi}}^{+}{\ensuremath{\mu}}^{\ensuremath{-}}{\ensuremath{\mu}}^{+}$ decay and we perform a fit to provide a first approach to it. We obtain ${\ensuremath{\delta}}_{\ensuremath{\rho}\ensuremath{\phi}}=(0.44\ifmmode\pm\else\textpm\fi{}0.24)\ensuremath{\pi}$. In contradistinction, it would be useful to consider observables insensitive to this phase. We first compute the invariant mass of the lepton pair at a given angle of one of the leptons emission with respect to the pion, and then we compute the forward-backward distribution at such angle. We show that for the latter observable the resonance region is smeared, exhibiting no dependence on the ${\ensuremath{\delta}}_{\ensuremath{\rho}\ensuremath{\phi}}$ phase, although global effects are observed by comparing with the pure phase space approach. In order to consider this observable in more general scenarios, we analyze the behavior for the resonant background process ${D}_{s}^{+}\ensuremath{\rightarrow}\ensuremath{\pi}\ensuremath{\pi}\ensuremath{\pi}$, the short distance contribution in the standard model, and the charged Higgs mediated process as given by the two Higgs doublet model, which exhibit distinguishable features among themselves.

Perceptions of Portuguese Doctors Regarding Hastened Death Scenarios: A Cross-Sectional Study.

A growing number of countries have legalized the process of hastening death. At a time when laws decriminalizing hastened death have been passed in the Portuguese Parliament, the development of research related with decision making regarding this issue is of crucial importance. This study seeks to evaluate, in a sample of Portuguese doctors, whether the presentation of clinical vignettes changes the agreement with the practice of hastened death compared with general scenarios. A questionnaire was distributed among academic physicians from medical schools across Portugal to assess their level of agreement or disagreement with the practice of hastened death. The questionnaire included eight standard cases and eight clinical vignettes framed under conditions defined by law for the practice of hastened death. Differences were analyzed using the t-Student test for paired samples. There were statistically significant differences in five scenarios (t = 3.46; p < 0.05; t = 2.47; p < 0.05; t = 4.28; p < 0.05; t = 3.38; p < 0.05; t = 3.66; p < 0.05) with greater agreement concerning the clinical vignettes. The highest acceptance was found in the requests made by adults with terminal and incurable illnesses. Agreement increased when the clinical vignette was presented in comparison with the respective standard for most of the cases of hastened death presented.