Discriminative graph convolution networks for hyperspectral image classification

Recently, the proposal of graph convolutional networks (GCN) has successfully implemented into hyperspectral image data representation and analysis. In spite of the great success, there are still several major challenges in hyperspectral image classification, including within-class diversity, and between-class similarity, which generally degenerate hyperspectral image classification performance. To address the problems, we propose a discriminative graph convolution networks (DGCN) for hyperspectral image classification. This method introduces the concepts of within-class scatter and between-class scatter, which respectively reflect the global geometric structure and discriminative information of the input space. The experimental results on the hyperspectral data sets show that the proposed method has good classification performance.     

Multi carrier energy systems and energy hubs: Comprehensive review,survey and recommendations

In this paper, the multi carrier energy (MCE) systems are reviewed from different point of views including mathematical models, integrated components and technologies, uncertainty management, planning objectives, environmental pollution, resilience, and robustness. The basic of MCE systems is formed by combination of cooling, heating and power (CCHP). The natural gas and electricity are the main inputs to MCE systems and the cooling, heating, and electricity are the common outputs. The regular energy converters in the MCE systems are combined heat and power (CHP), gas boiler, absorption-electrical chillers, power to gas (P2G) and fuel-cell. The generic energy storages are electrical, heating, cooling, hydrogen, carbon dioxide (CO₂) and hydro systems.     

基于GIS的航电枢纽安全监测信息管理系统设计与应用

为解决航电枢纽安全监测海量数据统筹管理以及GIS技术结合应用程度低等问题,开发了基于GIS的航电枢纽安全监测信息管理系统。倾斜摄影技术的应用实现了高效建立工程三维数字模型,通过实景模型与工程监测的结合,为监测点可视化管理及成果可视化展示提供了条件。系统实现的技术路线,可为类似工程项目提供参考。     

Assessing supply chain robustness to links failure

Supply chain networks need to respond efficiently to operation disruptions, as one of their aims is to guarantee the on time delivery of products. Hence, robustness has become one of the important issues to consider when designing supply networks. There are alternative ways to measure what robustness means in this context. In this paper, we propose a new metric based on the effect on service level of the collapse of active transportation links. Numerical experiments are carried out to understand how different design factors affect robustness. Robustness under a targeted attack is compared with robustness to random failures. Results show that flow complexity (i.e. the number of potential transportation links between supply network nodes) is the most influential factor affecting supply network and its robustness, as well as the service level that can be maintained after disruptions. Thus, diversification both in supply sources and transportation routes seems to be key to robustness.     

Consumer perception of extruded snacks containing brown rice and dried mushroom

Effects of mushroom type, seasoning and health benefit information (HBI) on consumers’ saltiness expectation, sensory liking, elicited emotions and purchase intent (PI) of extruded snacks were investigated. Five snacks were evaluated: straw mushroom (Volvariella volvacea) extrudates without (SME) or with seasoning (SMES), phoenix mushroom (Pleurotus pulmonarius) extrudates without (PME) or with seasoning (PMES), and the control without mushroom and seasoning. Hedonic scores and positive emotions were generally higher for seasoned mushroom-containing snacks (SMES and PMES) with 65% and 75.83% of consumers reporting willingness to purchase, respectively, after receiving HBI. Bored, interested and satisfied were identified as significant emotional predictors for PI odds. Flavour, saltiness, overall liking, bored, good and interested were critical attributes, differentiating snacks. This study demonstrated that sensory liking and PI of extruded brown rice-based snacks containing mushroom could be improved through savoury seasoning addition, which also allowed saltiness expectations to be met.     

Style-Controllable Speech-Driven Gesture Synthesis Using Normalising Flows

Automatic synthesis of realistic gestures promises to transform the fields of animation, avatars and communicative agents. In off-line applications, novel tools can alter the role of an animator to that of a director, who provides only high-level input for the desired animation; a learned network then translates these instructions into an appropriate sequence of body poses. In interactive scenarios, systems for generating natural animations on the fly are key to achieving believable and relatable characters. In this paper we address some of the core issues towards these ends. By adapting a deep learning-based motion synthesis method called MoGlow, we propose a new generative model for generating state-of-the-art realistic speech-driven gesticulation. Owing to the probabilistic nature of the approach, our model can produce a battery of different, yet plausible, gestures given the same input speech signal. Just like humans, this gives a rich natural variation of motion. We additionally demonstrate the ability to exert directorial control over the output style, such as gesture level, speed, symmetry and spacial extent. Such control can be leveraged to convey a desired character personality or mood. We achieve all this without any manual annotation of the data. User studies evaluating upper-body gesticulation confirm that the generated motions are natural and well match the input speech. Our method scores above all prior systems and baselines on these measures, and comes close to the ratings of the original recorded motions. We furthermore find that we can accurately control gesticulation styles without unnecessarily compromising perceived naturalness. Finally, we also demonstrate an application of the same method to full-body gesticulation, including the synthesis of stepping motion and stance.     

Passivity and synchronisation of complex dynamical networks with multiple derivative couplings

Two network models with multiple derivative couplings and different dimensions of output and input vectors are investigated in this paper. The problem of passivity for the proposed network models is analysed by utilising some inequality techniques and Lyapunov functional method, and several synchronisation conditions for complex dynamical networks with multiple derivative couplings (CDNMDC) are given. Moreover, by employing adaptive state feedback control strategy, some sufficient conditions for guaranteeing passivity and synchronisation of CDNMDC are obtained. In the end, we give two examples to verify the effectiveness of the results.     

An automated patient monitoring using discrete‐time wireless sensor networks

In recent years, Internet of Things (IoT) devices are used for remote health monitoring. For remotely monitoring a patient, only the health information at different time points are not sufficient; predicted values of biomarkers (for some future time points) are also important. In this article, we propose a powerful statistical model for an efficient dynamic patient monitoring using wireless sensor nodes through Bayesian Learning (BL). We consider the setting where a set of correlated biomarkers are measured from a patient through wireless sensors, but the sensors only report the ordinal outcomes (say, good, fair, high, or very high) to the sink based on some prefixed thresholds. The challenge is to use the ordinal outcomes for monitoring and predicting the health status of the patient under consideration. We propose a linear mixed model where interbiomarker correlations and intrabiomarker dependence are modeled simultaneously. The estimated and the predicted values of the biomarkers are transferred over the internet so that health care providers and the family members of the patient can remotely monitor the patient. Extensive simulation studies are performed to assess practical usefulness of our proposed joint model, and the performance of the proposed joint model is compared to that of some other traditional models used in the literature.     

Magnetic induction based cluster optimization in non-conventional WSNs: A cross layer approach

Conventional Radio Frequency (RF) communication technique is unsuitable for communication in non-conventional media (water, soil, rocks, etc.) because of heavy losses incurred due to dynamic channel characteristics. Magnetic Induction (MI) communication overcomes these losses as it is least affected by such varying channel characteristics. In non-conventional media based Wireless Sensor Networks (WSNs) the deployed sensor nodes cannot replace or replenish their batteries. Thus, energy consumption should be minimized and that can be achieved by clustering process. This process involves data sensing, aggregation and routing to the base station. These sub-tasks are performed under Physical (PHY), Medium Access Control (MAC) and Network (NET) layers of OSI Network model. Having lesser or larger number of clusters has different impact on energy consumption in different layers’ perspective. A large number of clusters decreases energy consumption as per PHY layer whereas it results in increased energy consumption as per MAC and NET layers. Thus, a trade-off is required to minimize the overall energy consumption. To this end, we found an optimal number of clusters considering the simultaneous influence of all three layers. The above analysis is performed for three media viz. sea water, fresh water and dry soil.     

The Spatial Patterning of Alcohol Outlets in Post-Disaster Christchurch,New Zealand: Practical and Policy Implications

Natural disasters greatly impact the environment of affected societies with often unknown consequences. In this study we examine the impact that the Canterbury Earthquakes of 2010–2011 had on the distribution of alcohol outlets in Christchurch, New Zealand. Specifically, we compare the distribution of both on-site and off-site alcohol outlets pre- (December 2009) and post-earthquake (December 2014) and use spatial regression models to identify whether any neighbourhood-level factors predict the observed redistributions. Overall, the number of alcohol outlets decreased by almost 13% after the Canterbury Earthquakes. However, we found a moderate increase in the clustering of both outlet types of outlets in the post-quake period. Increases in rates of both on-site and off-site alcohol outlets in neighbourhoods were positively associated with the percentage of residents who resided in their neighbourhood < 5 years and with neighbourhood crime rate change, while negative associations were found with percentage population aged between 15 and 29 years. The results suggest that the changing spatial distribution of alcohol outlets in Christchurch was not random but driven, in part, by the emergent demographic composition of neighbourhoods. The significant practical and policy implications of a redistribution of alcohol outlets are outlined providing a tangible link between empirical research and practice in an urban environment plagued with post-disaster spatial and social uncertainties.