Crowd Size Research Articles

A single-task and multi-decision evolutionary game model based on multi-agent reinforcement learning

In the evolutionary game of the same task for groups, the changes in game rules, personal interests, the crowd size, and external supervision cause uncertain effects on individual decision-making and game results. In the Markov decision framework, a single-task multi-decision evolutionary game model based on multi-agent reinforcement learning is proposed to explore the evolutionary rules in the process of a game. The model can improve the result of a evolutionary game and facilitate the completion of the task. First, based on the multi-agent theory, to solve the existing problems in the original model, a negative feedback tax penalty mechanism is proposed to guide the strategy selection of individuals in the group. In addition, in order to evaluate the evolutionary game results of the group in the model, a calculation method of the group intelligence level is defined. Secondly, the Q-learning algorithm is used to improve the guiding effect of the negative feedback tax penalty mechanism. In the model, the selection strategy of the Q-learning algorithm is improved and a bounded rationality evolutionary game strategy is proposed based on the rule of evolutionary games and the consideration of the bounded rationality of individuals. Finally, simulation results show that the proposed model can effectively guide individuals to choose cooperation strategies which are beneficial to task completion and stability under different negative feedback factor values and different group sizes, so as to improve the group intelligence level.

Understanding Season Ticket Holder Attendance Decisions

Predicting attendance at events is important for efficient facility management and marketing to maximize crowds. Most work to date is conducted at the aggregate level; however, the large crowd size being predicted often means important individual decisions are masked. In many markets, increased nonattendance by season ticket holders (STHs) is being reported, which is troubling given they have prepaid and are expected to be highly loyal. To understand who attends, rather than just how many, the authors analyze the “no-show” behavior of over 5,900 individual STH of one professional team over a season. Results show that in addition to game viewing and quality conditions, age, tenure, expenditure, and prior game attendance are predictors of individual attendance decisions, with differences in how individuals are influenced by winning and uncertainty of outcome. The paper expands understanding of drivers of STH attendance decisions and provides guidance toward managerial strategies for STH management.

Urban-Rural Compliance Variability to COVID-19 Restrictions of Indigenous Fijian (iTaukei) Funerals in Fiji.

Research on coronavirus disease 2019 (COVID-19) has focused primarily on impacts in Western societies despite emerging evidence of increased vulnerability among indigenous peoples such as Pacific Islanders. Using Facebook public posts, this research assessed compliance to COVID-19 restrictions such as social gatherings (SG) and social distancing (SD) in non-COVID-19, indigenous Fijian (iTaukei) funerals in Fiji. Results showed 95% of the 20 funerals exceeding SG limits with greater, and highly variable crowd sizes in rural than urban communities. Additionally, 75% of the 20 funerals did not adhere to the 2-m SD requirement of which 80% were in rural areas. Higher SG and SD compliance in urban funerals could be partially explained by the presence of a recognized authority who enforced crowd size limits, and the heterogeneous urban community who were more likely to flag breaches than their collectivistic, homogeneous, close-knit, rural counterparts. Ultimately, health authorities need to utilize a social lens that incorporates etic and emic differences in culture to ensure maximum compliance.

Crowd counting: a behavioural economics perspective.

Developments in technology have facilitated the emergence of new crowd counting organisations. Some of the organisations have established platforms to disseminate their data, making it available to researchers for the first time. These databases promise to increase the quality and quantity of research in various fields. In the late 2010s, specialist crowd counting organisations emerged with the sole purpose of counting crowds at protests and disseminating the results, sometimes in a purely partisan manner. Because of the contemporary relevance of protest behaviour, we frame our discussion within this context. For social scientists considering the utilisation of these new databases, it is essential that crowd numbers be linked to underlying human behaviour in a way that promises a chain of connections to investigate and explore. We use behavioural economics to show why relative crowd size may be important for human decision-makers. And we show how the significance of relative crowd size relates to other aspects of the human decision-making process, including risk preferences and probability assessments. Far from being a theory of protest behaviour, we present a behavioural economics-based primer for empirical researchers and social scientists engaging with newly available crowd counting data. The conclusions may apply in other contexts and might be extended to encompass specific types of behaviour, including aggression and violence. Indeed, the conclusions may guide the analysis of the emergence of the crowd counting organisations themselves.

Device-Free Crowd Counting Using Multi-Link Wi-Fi CSI Descriptors in Doppler Spectrum

Measuring the quantity of people in a given space has many applications, ranging from marketing to safety. A family of novel approaches to measuring crowd size relies on inexpensive Wi-Fi equipment, taking advantage of the fact that Wi-Fi signals get distorted by people’s presence, so by identifying these distortion patterns, we can estimate the number of people in such a given space. In this work, we refine methods that leverage Channel State Information (CSI), which is used to train a classifier that estimates the number of people placed between a Wi-Fi transmitter and a receiver, and we show that the available multi-link information allows us to achieve substantially better results than state-of-the-art single link or averaging approaches, that is, those that take the average of the information of all channels instead of taking them individually. We show experimentally how the addition of each of the multiple links information helps to improve the accuracy of the prediction from 44% with one link to 99% with 6 links.

Exit choice during evacuation is influenced by both the size and proportion of the egressing crowd

It is unclear how building occupants use information about the social and built environment when choosing an egress route during emergency evacuation. Conflicting tendencies have been previously reported: to follow the crowd, to avoid congestion, and to avoid unknown egress routes alone. We hypothesize that these tendencies depend on an interaction between social influence and the affordances (opportunities for egress) of the built environment. In three virtual reality (VR) experiments (each N=15), we investigated how social influence interacts with the affordances of available exits to determine exit choice. Participants were immersed in a crowd of virtual humans walking to the left or right exit and were asked to walk to one of the exits. Experiment 1 tested the role of social influence by manipulating both the proportion of the crowd walking toward one exit point (Crowd Proportion of 0 to 100%, in 10% increments) and the absolute number of virtual humans going to the exit point (Crowd Size of 10 or 20). Experiment 2 tested the role of affordances by introducing two visible exit doors (1 m width) in a closed room, and following the same protocol. Experiment 3 tested larger exit doors (2 m width) that afford rapid egress for more people. In the small crowd, participants were increasingly likely to follow the majority as its proportion increased. In the large crowd, however, participants tended to avoid the more crowded exit if the doors were narrow (Experiment 2), but not if the doors were wide (Experiment 3). Participants tended to follow a 100% majority in all experiments, thereby avoiding going to an exit alone. We propose that the dynamics of exit choice can be understood in terms of competition between alternative egress routes: the attraction of an exit increases with the proportion of the crowd moving toward it, becoming dominant at 100%, but decreases with the absolute number in the crowd moving toward it, relative to the exit’s affordance for egress.

Counting people in the crowd using social media images for crowd management in city events

City events are getting popular and are attracting a large number of people. This increase needs for methods and tools to provide stakeholders with crowd size information for crowd management purposes. Previous works proposed a large number of methods to count the crowd using different data in various contexts, but no methods proposed using social media images in city events and no datasets exist to evaluate the effectiveness of these methods. In this study we investigate how social media images can be used to estimate the crowd size in city events. We construct a social media dataset, compare the effectiveness of face recognition, object recognition, and cascaded methods for crowd size estimation, and investigate the impact of image characteristics on the performance of selected methods. Results show that object recognition based methods, reach the highest accuracy in estimating the crowd size using social media images in city events. We also found that face recognition and object recognition methods are more suitable to estimate the crowd size for social media images which are taken in parallel view, with selfies covering people in full face and in which the persons in the background have the same distance to the camera. However, cascaded methods are more suitable for images taken from top view with gatherings distributed in gradient. The created social media dataset is essential for selecting image characteristics and evaluating the accuracy of people counting methods in an urban event context.

Feddersen and Pesendorfer meet Ellsberg

The Condorcet Jury Theorem formalises the “wisdom of crowds”: binary decisions made by majority vote are asymptotically correct as the number of voters tends to infinity. This classical result assumes like-minded, expected utility maximising voters who all share a common prior belief about the right decision. Ellis (Theor Econo 11(3): 865–895, 2016) shows that when voters have ambiguous prior beliefs—a (closed, convex) set of priors—and follow maxmin expected utility (MEU), such wisdom requires that voters’ beliefs satisfy a “disjoint posteriors” condition: different private signals lead to posterior sets with disjoint interiors. Both the original theorem and Ellis’s generalisation assume symmetric penalties for wrong decisions. If, as in the jury context, errors attract asymmetric penalties then it is natural to consider voting rules that raise the hurdle for the decision carrying the heavier penalty for error (such as conviction in jury trials). In a classical model, Feddersen and Pesendorfer (Am Politi Sci Rev 92(1):23–35, 1998) have shown that, paradoxically, raising this hurdle may actually increase the likelihood of the more serious error. In particular, crowds are not wise under the unanimity rule: the probability of the more serious error does not vanish as the crowd size tends to infinity. We show that this “Jury Paradox” persists in the presence of ambiguity, whether or not juror beliefs satisfy Ellis’s “disjoint posteriors” condition. We also characterise the strictly mixed equilibria of this model and study their properties. Such equilibria cannot exist in the absence of ambiguity but may exist for arbitrarily large jury size when ambiguity is present. In addition to uninformative strictly mixed equilibria, analogous to those exhibited by Ellis (Theor Econo 11(3): 865–895, 2016), there may also exist strictly mixed equilibria which are informative about voter signals.

What Types of Crowds Generate More Valuable Content? Evidence from Cross-Platform Posting

This study examines the value of content generated by various groups of users on a financial social media platform. The value of such content is measured by the incremental accuracy of predicting stock volatility based on the cues from those content. We argue that the characteristics of a crowd such as crowd size, crowd diversity, and crowd independence have significant impact on the predictive value of the crowd-generated content. Leveraging a natural experiment setup where the financial platform no longer receives cross-postings from another major social media platform, we show empirical evidence that the composition of crowds (i.e., characteristics of the users who make up crowds) does matter. Furthermore, the impacts of the characteristic features on the value of crowd-generated content are likely heterogeneous and non-monotone. Finally, we discuss the implications of this study about digital platform envelopment and competition.

Is a large award truly attractive to solvers? The impact of award size on crowd size in innovation contests

AbstractFirms increasingly use human crowds to solve their innovation related needs. Innovation sourcing managers who sponsor innovation contests seek to maximize crowd size to increase the chance of creative outcomes. One strategy they use is to set an award size. Although many studies have examined the influence of award size on crowd size in innovation contests, the empirical findings are inconsistent and suggest further empirical study is needed. Drawing from expectancy‐value theory, we propose that award size and crowd size are related to one another in the form of an inverted U‐shape, namely that moderate awards maximize crowd size. Additional propositions cover contingencies around the number of awards and task difficulty of an innovation contest. Based on data representing 5342 programming innovation contests, our empirical analyses confirm the inverted U‐shaped relationship and indicate that the number of awards and task difficulty steepen the inverted U‐shaped relationship. These results add nuanced, theoretical understanding of how innovation contests work. The findings indicate that innovation sourcing managers should set award sizes at moderate levels, and that they should experiment with slightly larger or smaller awards to see if it impacts crowd size in their context.

Do zoo visitors induce attentional bias effects in primates completing cognitive tasks?

While previous research has focused on the impact of visitors on zoo-housed animals' behavior, here, we evaluated the impact of visitors on the performance of four zoo-housed Japanese macaques (Macaca fuscata) in a cognitive task. The macaques completed a touchscreen-based match-to-sample task in glass-sided booths at the perimeter of their enclosure, adjacent to a visitor viewing area. The task was novel to all macaques at the start of this study but over the 6-month testing period the macaques showed increased accuracy on the task, suggestive of learning. We recorded the number of visitors within the viewing area roughly every 12 trials each macaque completed. We categorized visitor counts as small (0-20), medium (21-40), and large (41-60) crowds and we considered the macaques' response latencies and accuracy by crowd size and study period (first 3months versus second 3months). If visitor presence negatively influenced performance, we predicted that macaques' accuracy would decrease but response times would increase with crowd size. We found effects of crowd size and study period on the macaques' accuracy. In the first period, the macaques performed at chance and accuracy did not differ across crowd categories. In the second period, the macaques' accuracy improved as compared to the first period, but their accuracy was mediated by crowd size: the macaques were significantly more accurate in the presence of small crowds than medium or large crowds. The macaques' response latencies also varied by study period and crowd size, but we found no evidence of a response-slowing effect.

Crowd Size Estimation Using CommSense Instrument for COVID-19 Echo Period

The period after the COVID-19 wave is called the Echo-period. Estimation of crowd size in an outdoor environment is essential in the Echo-period. Making a simple and flexible working system for the same is the need of the hour. This article proposes and evaluates a nonintrusive, passive, and cost-effective solution for crowd size estimation in an outdoor environment. We call the proposed system as LTE communication infrastructure based environment sensing or LTE-CommSense. This system does not need any active signal transmission as it uses LTE transmitted signal. So, this is a power-efficient, simple, low-footprint device. Importantly, the personal identity of the people in the crowd cannot be obtained using this method. First, the system uses practical data to determine whether the outdoor environment is empty or not. If not, it tries to estimate the number of people occupying the near range locality. Performance evaluation with practical data confirms the feasibility of this proposed approach.

Fire at the Oakland Ghost Ship Warehouse: A disaster life cycle-based analysis.

Structural fires remain a prominent threat to public health and safety even in several regions of the developed world, where rising housing costs force many to reside in unsafe environments. This case report of the Ghost Ship Warehouse fire in Oakland, California, highlights deficiencies in the emergency management system in the context of similar nightclub incidents to inform recommendations that might prevent such events from occurring in the future. The characteristics of the Ghost Ship warehouse and circumstances surrounding the fire, as described in government documents and news media sources, were examined using the disaster life cycle framework. The Ghost Ship fire was also compared with two prior fire disasters at the Happy Land nightclub in New York City and Station nightclub in West Warwick, Rhode Island. The following risk factors were identified as common features of deadly nightclub fires: large crowd size, limited access to exits, multiple code violations, lack of required permits, inadequate fire suppression systems, and poor building maintenance. To prevent the recurrence of such disasters, Oakland and other cities should adopt measures to improve interdepartmental communication; streamline reporting of fire and safety hazards, and allocate sufficient resources and staff capable of identifying hazardous buildings, performing inspections, and enforcing building and fire codes. Equally importantly, the urban affordable housing crisis needs to be adequately addressed to mitigate vulnerable populations living in unsafe dwellings that place them at risk of deadly fires.

Prediction modelling studies for medical usage rates in mass gatherings: A systematic review.

Mass gathering manifestations attended by large crowds are an increasingly common feature of society. In parallel, an increased number of studies have been conducted that developed and/or validated a model to predict medical usage rates at these manifestations. To conduct a systematic review to screen, analyse and critically appraise those studies that developed or validated a multivariable statistical model to predict medical usage rates at mass gatherings. To identify those biomedical, psychosocial and environmental predictors that are associated with increased medical usage rates and to summarise the predictive performance of the models. We searched for relevant prediction modelling studies in six databases. The predictors from multivariable regression models were listed for each medical usage rate outcome (i.e. patient presentation rate (PPR), transfer to hospital rate (TTHR) and the incidence of new injuries). The GRADE methodology (Grades of Recommendation, Assessment, Development and Evaluation) was used to assess the certainty of evidence. We identified 7,036 references and finally included 16 prediction models which were developed (n = 13) or validated (n = 3) in the USA (n = 8), Australia (n = 4), Japan (n = 1), Singapore (n = 1), South Africa (n = 1) and The Netherlands (n = 1), with a combined audience of >48 million people in >1700 mass gatherings. Variables to predict medical usage rates were biomedical (i.e. age, gender, level of competition, training characteristics and type of injury) and environmental predictors (i.e. crowd size, accommodation, weather, free water availability, time of the manifestation and type of the manifestation) (low-certainty evidence). Evidence from 3 studies indicated that using Arbon's or Zeitz' model in other contexts significantly over- or underestimated medical usage rates (from 22% overestimation to 81% underestimation). This systematic review identified multivariable models with biomedical and environmental predictors for medical usage rates at mass gatherings. Since the overall certainty of the evidence is low and the predictive performance is generally poor, proper development and validation of a context-specific model is recommended.

Large-Scale Dataset for Radio Frequency-Based Device-Free Crowd Estimation

Organisers of events attracting many people have the important task to ensure the safety of the crowd on their venue premises. Measuring the size of the crowd is a critical first step, but often challenging because of occlusions, noise and the dynamics of the crowd. We have been working on a passive Radio Frequency (RF) sensing technique for crowd size estimation, and we now present three datasets of measurements collected at the Tomorrowland music festival in environments containing thousands of people. All datasets have reference data, either based on payment transactions or an access control system, and we provide an example analysis script. We hope that future analyses can lead to an added value for crowd safety experts.

Fans and Match Results: Evidence From a Natural Experiment in Brazil

This paper assesses the effect of the crowd size on home team advantage, using a natural experiment occurred in Brazilian football. In 2015, the Brazilian Football Confederation (CBF) quasi-randomly assigned games to be played under an alternative schedule on Sundays at 11 a.m. This measure significantly increased attendance in assigned games. Using this alternative schedule as an instrumental variable, we show that the exogenous increase in attendance did not increase home team advantage.

Device-free crowd counting with WiFi channel state information and deep neural networks

Crowd counting is of great importance to many applications. Conventional vision-based approaches require line of sight and pose privacy concerns, while most radio-based approaches involve high deployment cost. In this paper, we propose to utilize WiFi channel state information (CSI) to infer crowd count in a device-free way, with only one pair of WiFi transmitter and receiver. The proposed method establishes the statistical relationship between the variation of CSI and the number of people with deep neural networks (DNN) and thereafter estimates the people count according to the real-time CSI through the trained DNN model. Evaluations demonstrate the effectiveness of the method. For the crowd size of 6, the counting error was within 1 person for 100% of the cases. For the crowd size of 34, the counting error was within 1 person for 97.7% of the cases and within 2 persons for 99.3% of the cases.

Robot Navigation in Crowds by Graph Convolutional Networks With Attention Learned From Human Gaze

Safe and efficient crowd navigation for mobile robot is a crucial yet challenging task. Previous work has shown the power of deep reinforcement learning frameworks to train efficient policies. However, their performance deteriorates when the crowd size grows. We suggest that this can be addressed by enabling the network to identify and pay attention to the humans in the crowd that are most critical to navigation. We propose a novel network utilizing a graph representation to learn the policy. We first train a graph convolutional network based on human gaze data that accurately predicts human attention to different agents in the crowd. Then we incorporate the learned attention into a graph-based reinforcement learning architecture. The proposed attention mechanism enables the assignment of meaningful weightings to the neighbors of the robot, and has the additional benefit of interpretability. Experiments on real-world dense pedestrian datasets with various crowd sizes demonstrate that our model outperforms state-of-art methods by 18.4% in task accomplishment and by 16.4% in time efficiency.

Pedestrian evacuation within limited-space buildings based on different exit design schemes

Considering the differences in the individual characteristics of pedestrians and the influence factors of buildings, we proposed a safety evacuation model for limited spaces. Evacuation efficiency, bottleneck area density, escape route characteristics, and similar factors were analyzed on the basis of different exit widths and operating doors. Results indicated that at an exit width of 1.1 m, the evacuation time was at a low-rising shock steady state with a normal crowd size. Therefore, the exit width of 1.1 m was the most advantageous in terms of evacuation time and construction cost. The aisle center was used more than the sides. The movement of pedestrians to the middle position affected the traffic capacity of the aisle. In comparison with outward-opening doors, inward-opening doors had a definite advantage in terms of evacuation efficiency, area density, and space utilization. When inward-opening doors are used, the escape path is consistent with the exit direction. The aisle width was reduced under the outward-opening door condition, and a bottleneck was formed. The escape route of pedestrians under the outward-opening door condition resembled the letter “Z”. Our experiment did not directly test the behavior in real evacuations but emphasized the role of different designs in real human decision making in a virtual environment. Our findings may be useful in identifying topics for future studies on real human crowd movement, exit design, and facility layout.

Performance indicators during international rugby union matches are influenced by a combination of physiological and contextual variables

ObjectivesResearch has linked physiological (e.g., hormonal, affective, fatigue) outcomes to performance indicators in rugby competition, but no work has integrated and contextualised these factors within a test-match environment. We addressed this gap by monitoring 29 athletes from a training squad across eight international rugby matches. DesignLongitudinal observational study. MethodsPre-match (8–9am) measures of salivary testosterone and cortisol concentrations, sleep duration, pulse rate, muscle soreness, stress, mood, and motivation were taken. Contextual factors were playing time, internal training load (ITL), test-match experience, opponent ranking, and crowd size. Performance was indexed by coach and player ratings of performance (CRP, PRP) and quantitative metrics; offloads, turnovers, runs with ball in hand (RWB), tackles, passes, and defenders beaten (DFB). ResultsMorning cortisol, sleep and mood were positively related to CRP and PRP (standardised coefficient estimates from 0.17 to 0.22). Cortisol, sleep, stress, mood and motivation were associated with one (or more) of turnovers, RWB, tackles, passes and DFB (incidence rate ratio [IRR] from 0.74 to 1.40). Playing time was positively related to all quantitative performance indicators (IRR from 1.01 to 1.04) with ITL, opponent ranking, and crowd size predicting selected outputs (IRR from 0.89 to 1.15). The explanatory models varied (conditional R2=0.15–0.83) but were generally stronger with both physiological and contextual inputs. ConclusionsMultiple physiological and contextual factors appear to contribute to player performance in international rugby competition. Measurement of these factors may guide training and management practices, a potential practical consequence but also advancing understanding from marker to causal link.