Search

Counting everything, also named few-shot counting, requires a model to be able to count objects with any novel (unseen) category giving few exemplar boxes. However, the existing few-shot counting methods are sub-optimal due to weak feature representation, such as the correlation between the exemplar patch and query feature, and contextual dependencies in density map prediction. In this paper, we propose a very simple but effective method, CSTrans, consisting of a Correlation-guided Self-Activation (CSA) module and a Local Dependency Transformer (LDT) module, to mitigate the above two issues, respectively. The CSA utilizes the correlation map to activate the semantic features and suppress the noisy influence of the query features, aiming at mining the potential relation while enriching correlation representation. Furthermore, the LDT incorporates a Transformer to explore local contextual dependencies and predict the density map. Our method achieves competitive performance on FSC-147 and CARPK datasets. We hope its simple implementation and superior performance can serve as a new and strong baseline for few-shot counting tasks and attract more interest in designing simple but effective models in future studies. Our code for CSTrans is available at https://github.com/gaobb/CSTrans. • A simple but effective CSTrans framework for counting everything. • A correlation-guided self-activation module for enriching feature representation. • A local dependency transformer module for modeling local context dependency. • Excellent performance on two few-shot counting datasets, FSC-147and CARPK. [ABSTRACT FROM AUTHOR]

This paper presents a Cartesian grid based tailored finite point method (TFPM) for singularly perturbed reaction-diffusion equation on complex domains. The method is incorporated with the kernel-free boundary integral algorithm, where the semi-discrete boundary value problems after time integration are reformulated into corresponding Fredholm boundary integral equations (BIEs) of the second kind, however with no algorithmic dependence on the exact analytical expression for the kernels of integrals. The BIEs are iteratively solved by the GMRES method while integral evaluation during each iteration resorts to solving an equivalent interface problem, which in practice is achieved by a series of manipulations in the framework of TFPM including discretization, correction, solution, and interpolation. The proposed method has second-order accuracy for the reaction-diffusion equation as demonstrated by the numerical examples. [ABSTRACT FROM AUTHOR]

• A cooperation mode for cross-organizational construction resource sharing is proposed. • Quantitative methods are employed for cooperation alliance development. • Contract-obedient profit allocation and contract-violated compensation are considered. • Sustainable construction can be facilitated via inter-site formwork reuse. Concrete formwork reuse enables both cost savings and resource consumption reduction and has attracted more attention recently. Construction sites in a certain district can reuse idle or underutilized formwork via cooperation. However, the way to cooperate and allocate cooperation profits is rarely explored. In addition, existing efforts usually assume that there are no contract-violated members in the cooperation. This is not in line with the practice since on-site emergencies always result in construction changes. This study aims to explore an applicable cooperation mode to improve the feasibility and stability of inter-site formwork reuse. The proposed cooperation mode enables optimal cooperation alliance establishment, contract-obedient profit allocation, and contract-violated compensation distribution and is illustrated and demonstrated via a case study in Chongqing, China. The results show that (1) formwork reuse within the optimal cooperation alliance can effectively reduce both costs and material consumptions; (2) profit allocation can motivate construction sites to cooperate for formwork reuse; and (3) compensation and penalties can guarantee the deserved profits of contract-obedient alliance members and prevent a breach of contract. These findings show that the proposed cooperation mode can promote formwork reuse among construction sites and provide insights into the sustainable development of cross-organizational cooperation. [ABSTRACT FROM AUTHOR]

Voice alarms play an important role in the evacuation of metro systems, while few studies focus on voice alarms inside train carriages. In this study, the effect of a fire warning on the initial phase of evacuation in a carriage is investigated. Both a questionnaire investigation and evacuation experiments with different fire voice alarms in a metro train model were conducted to quantitatively analyze the behaviors of evacuees during the response phase of evacuation. According to the analysis of the questionnaire, the response time calculated from the danger signal (smoke) is suitable for assessing the evacuees' response during the response phase. As indicated by the Pearson correlation coefficient between the minimum and mean response times (0.96), the first person's response has a significant influence on the behaviors of other evacuees. Moreover, comparisons among different categories of experiments demonstrate that the danger signal (smoke), fire alarm bell, and 3–4 units of a voice alarm with a normal and comfortable tone can deliver effective information to evacuees during the initial phase of evacuation. In addition, the effect of a voice alarm with an urgent and fast tone still needs further investigation. The results can help design fire voice alarms and improve the efficiency of evacuation in metro systems. [ABSTRACT FROM AUTHOR]

Abstract: This work proposes a tailored finite point method (TFPM) for the numerical solution of an anisotropic diffusion problem, which has much smaller diffusion coefficient along one direction than the other on a rectangular domain. The paper includes analysis on the differentiability of the solution to the given problem under some compatibility conditions. It has detailed derivation for a semi-discrete TFPM for the given problem. This work also proves a uniform error estimate on the approximate solution. Numerical results show that the TFPM is accurate as well as efficient for the strongly anisotropic diffusion problem. Examples include those that do not satisfy compatibility and regularity conditions. For the incompatible problems, numerical experiments indicate that the method proposed can still offer good numerical approximations. [Copyright &y& Elsevier]

In this paper, the numerical solutions of the problems of heat equation in two dimensions on unbounded domains are considered. For a given problem, we introduce an artificial boundary Γ to finite the computational domain. On the artificial boundary Γ, we propose an exact boundary condition to reduce the given problem to an initial-boundary problem of heat equation on the finite computational domain, which is equivalent to the original problem. Furthermore, a series of approximating artificial boundary conditions is given. Then the finite difference method and finite element method are used to solve the reduced problem on the finite computational domain. Finally, the numerical examples show the feasibility and effectiveness of the method given in this paper. [ABSTRACT FROM AUTHOR]

In this paper, the numerical solutions of three problems of heat equation on unbounded domains are considered. For each problem, we introduce an artificial boundary Γ to make the computational domain finite. On the artificial boundary Γ, we propose an exact artificial boundary condition to reduce the original problem to an initial-boundary problem of heat equation on the finite computational domain, which is equivalent to the original problem. Then the finite difference method and finite element method are used to solve the reduced problem on the finite computational domain. In the end of this paper, three numerical examples show the feasibility and effectiveness of the method given in this paper. [Copyright &y& Elsevier]

• The evacuation process was divided into 4 stages based on five events. • A shorter voice alarm cycle, a closer voice source and a more rapid tone resulted in a shorter recognition time. • Individual response time followed a Weibull distribution and the response order was affected by physical distance. • A closer and more directive voice guided evacuees to find the exit more quickly. • Opening doors by using emergency devices took about 75% of the transfer time on average. Controlled experiments involved more than 700 participants inside a metro train model were conducted to investigate effects of voice alarms on the evacuation process. The process was divided into 4 stages from the perspective of collective behavior: the recognition stage (from the emergency happens to the first person standing up), the response stage (from the first person standing up to the last person standing up), the transfer stage (from the last person standing up to the first person leaving the carriage), and the queue stage (from the first person leaving the carriage to the last person leaving the carriage). Results showed that the mean evacuation time of all experiments was 77.61 s, and mean percentages of the four stages were 16.89%, 12.80%, 30.67% and 30.76%, respectively. When focusing on a specific stage, a shorter voice alarm cycle, a closer voice source and a more rapid tone resulted in a shorter recognition time. In the response stage, the response order was mainly affected by physical distance. A closer and more directive voice guided evacuees to find the exit more quickly, and further reduced the transfer time. Opening doors by using emergency devices took about 75% of the transfer time on average. Voice alarm didn't play a key role in the queue stage, and the queue time was mainly affected by the number and the width of available doors. The results provide references for developing models and designing Public Announcement system of a metro train. [ABSTRACT FROM AUTHOR]

• In this paper, we propose an iterative splitting method as a combination of operator splitting and iterative methods to solve option pricing problems. To guarantee convergence, we use a mixed method to improve the numerical scheme. And a sufficient condition of the convergence of the iteration is given. • We focus on the Heston model and construct artificial boundary conditions (ABCs) for the problem in addition to the boundary condition in Heston's original paper. • Numerical experiments show that the iterative splitting method gives significantly better solutions compared to the classic 2-dimensional finite difference scheme no matter which ABC is chosen. • The iterative splitting method has a big advantage as well when calculating the Greeks including Delta, Gamma and Vega which have received much attention in practice. • Our new method transforms a 2-dimenstional problem into quasi 1-dimensional ones so that it helps to reduce computational cost and can be easily extended to other models for option valuation. In this paper, we propose an iterative splitting method to solve the partial differential equation (PDE) in option pricing problems. We focus on the Heston stochastic volatility model and the derived two-dimensional PDE. We take the European option as an example and conduct numerical experiments using different boundary conditions. The iterative splitting method transforms the two-dimensional equation into two quasi one-dimensional equations with the variable on the other dimension fixed, which helps to lower the computational cost. Numerical results show that the iterative splitting method together with an artificial boundary condition (ABC) based on the method by Li and Huang (2019) gives the most accurate option price and Greeks compared to the classic finite difference method with the commonly-used boundary conditions in Heston (1993). [ABSTRACT FROM AUTHOR]

• An experiment on pedestrian turning movement on the stair landing is conducted. • Pedestrian descent and ascent behaviors on the stair landing are compared. • Two phases can be found based on pedestrian speed on the stair landing. • Two regimes can be spotted for fundamental diagrams on the stair landing. • Empirical values for parameters in the method based on hypothetical travel path are provided. In this paper, we conduct a trajectory-based analysis on pedestrian turning movement on a stair landing. The extracted trajectories indicate pedestrians walk along arc paths. The compression effect resulting from pedestrians' desire to seek the shortest path during the turning movement is not observed. Based on the developing trends of angular speed and speed, two phases can be distinguished during the whole turning process, namely deceleration and uniform speed phase when moving downstairs, and acceleration and uniform speed phase when moving upstairs. According to the fundamental diagrams, compared with ascent movement, pedestrians move faster when going downstairs on the stair landing, and the speed is much lower than that in the straight corridor, due to the effect of turning movement and stairs. Two regimes can be spotted from the fundamental diagrams. When the density is lower than the critical one, the speed decreases sharply with the increase of density, while it decreases smoothly when the density is higher than the critical one. According to the density distributions, the area with the maximal density occurs after turning for descent movement, while it occurs before turning for ascent movement. Taking the average speed computed from trajectories as a reference, we further provide some empirical values for parameters (e.g., the turning radius of the arc path) proposed in the method based on hypothetical travel path. [ABSTRACT FROM AUTHOR]

Nonreciprocal interaction crowd systems, such as human–human, human–vehicle, and human–robot systems, often have serious impacts on pedestrian safety and social order. A more comprehensive understanding of these systems is needed to optimize system stability and efficiency. Despite the importance of these interactions, empirical research in this area remains limited. Thus, in our study we explore this underresearched area, focusing on scenarios where nonreciprocity plays a critical role, such as mass stabbings, which pose a substantial risk to public safety. We conducted the first experiments on this system and analysed high-accuracy data obtained from these experiments. Specifically, we conduct laboratory experiments on three scenarios: single exits, dual exits, and obscured chases. Then, we introduce an equation to describe the mechanism behind a direct threat zone, which significantly affects pedestrian behaviour. The extent of the direct threat zone is determined by the speed of the moving threat and the radius of danger occurrence. The equation can provide insights for different nonreciprocal interaction crowd systems, where the subjects of the nonreciprocal interactions can be a pedestrian, an autonomous robot or a vehicle. We further categorize potential threats into direct, adjacent, and rear-view zones, quantifying the level of threat for pedestrians. Our study revealed that a pedestrian's desired velocity correlated positively with potential threat intensity, increasing until near the direct threat zone. An emerging steady state is observed when escape routes are blocked by moving threats. This deviation affects the density–velocity relationship, making it distinct from the general relationship. This deviation signifies unique pedestrian behaviour in the presence of moving threats. Additionally, the rate of change in the angle for pedestrian motion in various desired directions is synchronized. This indicates the emergence of collective intelligence in nonreciprocal interaction crowd systems. As a result, our study may constitute a pioneering step towards understanding nonreciprocal interactions in crowd systems through laboratory experiments. These findings may enhance pedestrian safety and inform not only government crowd management strategies but also individual self-protection measures. • First high-accuracy lab experiment on strongly coupled nonreciprocal crowd systems. • Modelling non-reciprocal pedestrian-attacker interactions through potential threats. • Threat increases nearby pedestrian's desired speed; relationship is nonlinear. • Collective intelligence emergence and its causes observed in experiments. [ABSTRACT FROM AUTHOR]

• GhIAA43, a member of cotton Aux/IAA, is responsive to both IAA treatment and V. dahliae infection. • Silencing GhIAA43 enhanced cotton resistance to Verticillium wilt. • Silencing GhIAA43 activates SA-related defenses in cotton. • Auxin signaling and SA production positively interplay in cotton- Verticillium dahliae pathosystem. Auxin-mediated degradation of Aux/IAA proteins is a crucial step in auxin signaling. Recent researches indicate that Aux/IAA members also play a role in biotic and abiotic stresses. For example, Pseudomonas syringae infection causes Arabidopsis Aux/IAA protein (AXR2, AXR3) turnover. Here, by analyzing RNA-seq data we found that several cotton Aux/IAA genes are responsive to Verticillium dahliae infection, one of these named GhIAA43 , was investigated for its role in cotton defense against V. dahliae infection. We demonstrate that the transcript levels of GhIAA43 were responsive to both V. dahliae infection and exogenous IAA application. By producing transgenic Arabidopsis plants overexpressing GhIAA43-GUS fusion, we show that IAA treatment and V. dahliae infection promoted GhIAA43 protein turnover. Silencing GhIAA43 in cotton enhanced wilt resistance, suggesting that GhIAA43 is a negative regulator in cotton defense against V. dahliae attack. By monitoring SA marker gene expression and measurement of SA content in GhIAA43 -silenced cotton plants, we found that the enhanced resistance in GhIAA43 -silenced cotton plants is due to the activation of SA-related defenses, and the activated defenses specifically occurred in the presence of V. dahliae. Furthermore, exogenous IAA application improve wilt resistance in cotton plants tested. Our results provide novel connection between auxin signaling and SA-related defenses in cotton upon V. dahliae attack. [ABSTRACT FROM AUTHOR]

• A seismic casualty prediction method that considers the rescue process is constructed. • In this method, a seismic network model with a dynamic rescue rate is built. • The optimal rescue plan can be obtained by solving the seismic network model. • The optimal rescue plan can minimize the casualties of earthquake. To predict the impact of the rescue work on the casualties after the earthquake, a seismic casualty prediction method that comprehensively considers the burial situation and the rescue process is constructed. In this method, a seismic network model based on the graph theory method is built, then the burial situation on an area of after an earthquake can be quickly calculated by using the calculation method of casualty rate based on the building damage rate. Finally, the optimal rescue plan can be obtained by solving the objective function of the seismic network model through the genetic algorithm. Taking the 1100 campus of University of Shanghai for Science and Technology as a case, the burial situation and the casualties under the influence of the rescue work are analyzed after assuming a Ⅹ-degree earthquake occurs on this area, based on which an optimal rescue plan is proposed. After that, by comparing with two shortest path methods, the superiority of the proposed method is highlighted. This method can be used as the basis for regional earthquake risk assessment and emergency rescue plans establishment, and it can also be used as a tool for generating optimal rescue routes after the earthquake. [ABSTRACT FROM AUTHOR]

The usage of efficient electrocatalyst in an appropriate way can significantly improve the oxygen reduction performance of microbial fuel cells (MFCs) cathode, which may accelerate its commercial application. In this work, isoreticular metal organic framework-3 (IRMOF-3) that modified with g-C 3 N 4 and pyrolyzed at 850 °C (IR/CN-x%), exhibits a superior catalytic activity in alkaline electrolytes for oxygen reduction reaction (ORR). At high temperatures, after zinc evaporation, IRMOF-3 converts to a metal-free carbon skeleton that has large specific surface area. As a nitrogen source and supporting template, g-C 3 N 4 introduces more activated nitrogen, and forms a new metal-free hierarchically porous nitrogen-doped carbon structure. As-synthesized IR/CN-50% as cathode catalyst displays a half-wave potential of 0.89 V (vs. RHE) and limit current density of 6.35 mA cm−2 in 0.1 M KOH, which is superior to commercial Pt/C (0.86 V@5.51 mA cm−2). Furthermore, IR/CN-50% shows approximately four-electron pathway and a yield of less than 4% hydrogen peroxide. Results from electrochemical tests further confirm IR/CN-50% possesses better durability and higher methanol tolerance than Pt/C. Besides, MFCs modified with IR/CN-50% present maximum power density (1402.8 mW m−3) better than the MFCs equipping with Pt/C catalyst (1292.8 mW m−3), which is a promising alternative electrocatalyst for MFCs. Image 1 • The g-C 3 N 4 as both nitrogen source and supporting template. • Metal-free nitrogen-doped interconnected porous carbon structure. • Abundant active sites and mesoporous structures. • Applying as the cathode catalyst in microbial fuel cells. [ABSTRACT FROM AUTHOR]

Despite its great promise in non-invasive treatment of cancers, magnetic resonance-guided focused ultrasound surgery (MRgFUS) is currently limited by the insensitivity of magnetic resonance imaging (MRI) for visualization of small tumors, low efficiency of in vivo ultrasonic energy deposition, and damage to surrounding tissues. We hereby report the development of an active targeting nano-sized theranostic superparamagnetic iron oxide (SPIO) platform for significantly increasing the imaging sensitivity and energy deposition efficiency using a clinical MRgFUS system. The surfaces of these PEGylated SPIO nanoparticles (NPs) were decorated with anti-EGFR (epidermal growth factor receptor) monoclonal antibodies (mAb) for targeted delivery to lung cancer with EGFR overexpression. The potential of these targeted nano-theranostic agents for MRI and MRgFUS ablation was evaluated in vitro and in vivo in a rat xenograft model of human lung cancer (H460). Compared with nontargeting PEGylated SPIO NPs, the anti-EGFR mAb targeted PEGylated SPIO NPs demonstrated better targeting capability to H460 tumor cells and greatly improved the MRI contrast at the tumor site. Meanwhile, this study showed that the targeting NPs, as synergistic agents, could significantly enhance the efficiency for in vivo ultrasonic energy deposition in MRgFUS. Moreover, we demonstrated that a series of MR methods including T2-weighted image (T2WI), T1-weighted image (T1WI), diffusion-weighted imaging (DWI) and contrast-enhanced T1WI imaging, could be utilized to noninvasively and conveniently monitor the therapeutic efficacy in rat models by MRgFUS. [ABSTRACT FROM AUTHOR]

• Subject repeat the movement from 12th floor to the ground up to 11 times. • A new point is proposed and verified, that is, as the expected movement distance increases, the movement speed of the person will decrease. • The relation between the increase rate of evacuation distance and the decrease rate of speed is presented. • With the increase of expected evacuation distance, the speed decrease rate of the male is greater than that for female. • Fatigue also has an important effect on the descent speed. Currently, there are an increasing number of super high-rise buildings in urban cities, and the issue of evacuation in emergencies from super-long evacuation distance in stairwells comes to the fore. An evacuation experiment was carried out in Shanghai Tower (Fang et al., 2020), it was found that the evacuation speed of pedestrians evacuated from the 126th floor was always slower than that of those from the 117th floor. Therefore, we propose a hypothesis that the expected evacuation distance will affect pedestrians' movement speed. Pedestrians will adjust the evacuation speed according to the expected evacuation distance, which may be caused by psychological expectation. In order to verify our conjecture, we conduct an experiment in a 12-storey office building, that is, to study whether there would be an influence and what kind of influence would be caused on speed by setting the evacuation distance for participants in advance. According to the results, we find that with the increase of expected evacuation distance, the movement speed of pedestrians will decrease, which confirms our hypothesis. The groups who completed 8 laps had the least evacuation distance, and its average speed is far faster than that of G_11 with longest evacuation distance. Affecting by psychological factors, the individual average descending speed of G_5 + 3 is all over 1.07 m/s in each lap of 8 laps and the maximum average speed is 1.30 m/s while G_11 has the lowest average individual descent speed (the maximum average speed is only 1.09 m/s) in 7 scenarios. Fatigue has an important effect on the descent speed, and the descent speed tends to decrease with the increase of evacuation distance. After the rest, the speed of male, female and all participants increased by 28.0%, 28.5% and 27.7%, respectively. At the same time, we give the relation between the increase rate of evacuation distance and the decrease rate of speed. It also can be found that with the increase of expected evacuation distance, the speed decrease rate of the male (the slope of the downward trend is −0.183) is greater than that for females (the slope of the downward trend is −0.138). In addition, we study the effects of actual evacuation distance, sex, BMI on evacuation speed. Finally, we obtain the correlation between heart rate and speed. The results in this paper are beneficial to the study of pedestrians' characteristics in the long and super-long distance evacuation. [ABSTRACT FROM AUTHOR]