Crowd Data Research Articles

Virtual Crowds Rheology: Evaluating the Effect of Character Representation on User Locomotion in Crowds.

Crowd data is a crucial element in the modeling of collective behaviors, and opens the way to simulation for their study or prediction. Given the difficulty of acquiring such data, virtual reality is useful for simplifying experimental processes and opening up new experimental opportunities. This comes at the cost of the need to assess the biases introduced by the use of this technology. Our paper is part of this effort, and investigates the effect of the graphical representation of a crowd on the behavior of a user immersed within. More specifically, we inspect the virtual navigation through virtual crowds, in terms of travel speeds and local navigation choices as a function of the visual representation of the virtual agents that make up the crowd (simple geometric model, anthropomorphic model or realistic model). Through an experiment in which we ask a user to navigate virtual crowds of varying densities, we show that the effect of the visual representation is limited, but that an anthropomorphic representation offers the best trade-off between computational complexity and ecological validity, even though a more realistic representation can be preferred when user behaviour is studied in more details. Our work leads to clear recommendations on the design of immersive simulations for the study of crowd behavior.

The Influence of Fixed and Flexible Funding Mechanisms on Reward-Based Crowdfunding Success

This study examined whether fixed or flexible funding mechanisms influence crowdfunding success. Under the fixed funding mechanism, the pledges contributed to the crowdfunding campaign projects are returned to the backers if the project fails, whereas, under the flexible funding mechanism, the project creator can keep all the raised pledges, irrespective of whether the project succeeds or fails. Secondary data consisted of reward-based crowdfunding projects retrieved from The Crowd Data Centre. Logistic regression was employed to respond to research objectives. The results reveal that the fixed funding mechanism increases the probability of success more than flexible funding. Entrepreneur experience, spelling errors, and project description negatively affect crowdfunding success, and backers positively affect crowdfunding success. The findings guide entrepreneurs seeking financing to design and choose an appropriate funding mechanism that effectively reduces the failure rate. Although many entrepreneurs seek funding in the crowdfunding market, relatively little research has been conducted on the influence of flexible or fixed funding mechanisms on crowdfunding success in Africa. This study provides entrepreneurs with appropriate financing strategies that enhance crowdfunding success. The empirical literature indicates that the flexible funding mechanism creates distrust among backers due to unrealistic target amounts.

Violence detection in crowd videos using nuanced facial expression analysis

Video analysis for violence detection is crucial, especially when dealing with crowd data, where the potential for severe mob attacks in sensitive areas is high. This paper proposes a solution utilizing Convolutional Restricted Boltzmann Machine (CRBM) for video analysis, integrating the strengths of Convolutional Neural Network (CNN) and Restricted Boltzmann Machine (RBM). By focusing on image patches rather than entire frames, the method addresses the challenge of object detection in crowded scenes. The CRBM combines deep-level image analysis from CNN with unsupervised feature extraction in RBM, facilitated by image convolution using Gabor filters in the hidden layer. Dropout regularization mitigates overfitting, enhancing model generality. Extracted features are inputted into an SVM classifier for face detection and a custom VGG16 model for emotion identification. Event probability is then determined through logistic regression based on facial expressions. Despite existing approaches for smart crowd behaviour identification, there remains a tradeoff between accuracy and processing time. Our proposed solution addresses this by employing proper frame preprocessing techniques for feature extraction. Validation using quantitative and qualitative metrics confirms the effectiveness of the approach.

With or Without You: Effect of Contextual and Responsive Crowds on VR-based Crowd Motion Capture.

While data is vital to better understand and model interactions within human crowds, capturing real crowd motions is extremely challenging. Virtual Reality (VR) demonstrated its potential to help, by immersing users into either simulated virtual crowds based on autonomous agents, or within motion-capture-based crowds. In the latter case, users' own captured motion can be used to progressively extend the size of the crowd, a paradigm called Record-and-Replay (2R). However, both approaches demonstrated several limitations which impact the quality of the acquired crowd data. In this paper, we propose the new concept of contextual crowds to leverage both crowd simulation and the 2R paradigm towards more consistent crowd data. We evaluate two different strategies to implement it, namely a Replace-Record-Replay (3R) paradigm where users are initially immersed into a simulated crowd whose agents are successively replaced by the user's captured-data, and a Replace-Record-Replay-Responsive (4R) paradigm where the pre-recorded agents are additionally endowed with responsive capabilities. These two paradigms are evaluated through two real-world-based scenarios replicated in VR. Our results suggest that the behaviors observed in VR users with surrounding agents from the beginning of the recording process are made much more natural, enabling 3R or 4R paradigms to improve the consistency of captured crowd datasets.

Domain-Adaptive Crowd Counting via High-Quality Image Translation and Density Reconstruction.

Recently, crowd counting using supervised learning achieves a remarkable improvement. Nevertheless, most counters rely on a large amount of manually labeled data. With the release of synthetic crowd data, a potential alternative is transferring knowledge from them to real data without any manual label. However, there is no method to effectively suppress domain gaps and output elaborate density maps during the transferring. To remedy the above problems, this article proposes a domain-adaptive crowd counting (DACC) framework, which consists of a high-quality image translation and density map reconstruction. To be specific, the former focuses on translating synthetic data to realistic images, which prompts the translation quality by segregating domain-shared/independent features and designing content-aware consistency loss. The latter aims at generating pseudo labels on real scenes to improve the prediction quality. Next, we retrain a final counter using these pseudo labels. Adaptation experiments on six real-world datasets demonstrate that the proposed method outperforms the state-of-the-art methods.

GREIL-Crowds: Crowd Simulation with Deep Reinforcement Learning and Examples

Simulating crowds with realistic behaviors is a difficult but very important task for a variety of applications. Quantifying how a person balances between different conflicting criteria such as goal seeking, collision avoidance and moving within a group is not intuitive, especially if we consider that behaviors differ largely between people. Inspired by recent advances in Deep Reinforcement Learning, we propose Guided REinforcement Learning (GREIL) Crowds, a method that learns a model for pedestrian behaviors which is guided by reference crowd data. The model successfully captures behaviors such as goal seeking, being part of consistent groups without the need to define explicit relationships and wandering around seemingly without a specific purpose. Two fundamental concepts are important in achieving these results: (a) the per agent state representation and (b) the reward function. The agent state is a temporal representation of the situation around each agent. The reward function is based on the idea that people try to move in situations/states in which they feel comfortable in. Therefore, in order for agents to stay in a comfortable state space, we first obtain a distribution of states extracted from real crowd data; then we evaluate states based on how much of an outlier they are compared to such a distribution. We demonstrate that our system can capture and simulate many complex and subtle crowd interactions in varied scenarios. Additionally, the proposed method generalizes to unseen situations, generates consistent behaviors and does not suffer from the limitations of other data-driven and reinforcement learning approaches.

Anomaly Prediction over Human Crowded Scenes via Associate-Based Data Mining and K-Ary Tree Hashing

Anomaly detection and behavioral recognition are key research areas widely used to improve human safety. However, in recent times, with the extensive use of surveillance systems and the substantial increase in the volume of recorded scenes, the conventional analysis of categorizing anomalous events has proven to be a difficult task. As a result, machine learning researchers require a smart surveillance system to detect anomalies. This research introduces a robust system for predicting pedestrian anomalies. First, we acquired the crowd data as input from two benchmark datasets (including Avenue and ADOC). Then, different denoising techniques (such as frame conversion, background subtraction, and RGB-to-binary image conversion) for unfiltered data are carried out. Second, texton segmentation is performed to identify human subjects from acquired denoised data. Third, we used Gaussian smoothing and crowd clustering to analyze the multiple subjects from the acquired data for further estimations. The next step is to perform feature extraction to multiple abstract cues from the data. These bag of features include periodic motion, shape autocorrelation, and motion direction flow. Then, the abstracted features are mapped into a single vector in order to apply data optimization and mining techniques. Next, we apply the associate-based mining approach for optimized feature selection. Finally, the resultant vector is served to the k-ary tree hashing classifier to track normal and abnormal activities in pedestrian crowded scenes.

Data-Driven Real-Time Denied Boarding Prediction in Urban Railway Systems

Providing real-time crowding information in urban railways would enable informed travel decisions and encourage cooperative behavior of passengers, as well as improve operating efficiency and safety. However, the problem of real-time crowding prediction is not trivial because of the unavailability of ground-truth crowding data, particularly for the direct impact of crowding on passengers (e.g., denied boarding on platforms). This paper proposes a data-driven method for real-time denied boarding prediction in urban railway systems using automated fare collection (AFC) and automated vehicle location (AVL) data. It predicts the denied boarding probability distribution or its derived metrics as a function of explanatory variables, including demand, operations, and incident-related factors. The method is validated through a case study covering 18 months on Hong Kong Mass Transit Railways. The results highlight the model’s accurate and robust performance in predicting denied boarding on platforms using purely AFC and AVL data (e.g., an average of 6%–7% error) under both recurrent and non-recurrent situations, in which transfer demand-related factors contribute most to the prediction. The prediction model outputs can support proactive operations control and management, as well as provision of customer information without relying on expensive and time-consuming data collection efforts. Customer information, for example, can include the average number of trains to wait before boarding or the average waiting time on the platform.

Model‐based Crowd Behaviours in Human‐solution Space

AbstractRealistic crowd simulation has been pursued for decades, but it still necessitates tedious human labour and a lot of trial and error. The majority of currently used crowd modelling is either empirical (model‐based) or data‐driven (model‐free). Model‐based methods cannot fit observed data precisely, whereas model‐free methods are limited by the availability/quality of data and are uninterpretable. In this paper, we aim at taking advantage of both model‐based and data‐driven approaches. In order to accomplish this, we propose a new simulation framework built on a physics‐based model that is designed to be data‐friendly. Both the general prior knowledge about crowds encoded by the physics‐based model and the specific real‐world crowd data at hand jointly influence the system dynamics. With a multi‐granularity physics‐based model, the framework combines microscopic and macroscopic motion control. Each simulation step is formulated as an energy optimization problem, where the minimizer is the desired crowd behaviour. In contrast to traditional optimization‐based methods which seek the theoretical minimizer, we designed an acceleration‐aware data‐driven scheme to compute the minimizer from real‐world data in order to achieve higher realism by parameterizing both velocity and acceleration. Experiments demonstrate that our method can produce crowd animations that are more realistically behaved in a variety of scales and scenarios when compared to the earlier methods.

Domain-General Crowd Counting in Unseen Scenarios

Domain shift across crowd data severely hinders crowd counting models to generalize to unseen scenarios. Although domain adaptive crowd counting approaches close this gap to a certain extent, they are still dependent on the target domain data to adapt (e.g. finetune) their models to the specific domain. In this paper, we instead target to train a model based on a single source domain which can generalize well on any unseen domain. This falls into the realm of domain generalization that remains unexplored in crowd counting. We first introduce a dynamic sub-domain division scheme which divides the source domain into multiple sub-domains such that we can initiate a meta-learning framework for domain generalization. The sub-domain division is dynamically refined during the meta-learning. Next, in order to disentangle domain-invariant information from domain-specific information in image features, we design the domain-invariant and -specific crowd memory modules to re-encode image features. Two types of losses, i.e. feature reconstruction and orthogonal losses, are devised to enable this disentanglement. Extensive experiments on several standard crowd counting benchmarks i.e. SHA, SHB, QNRF, and NWPU, show the strong generalizability of our method. Our code is available at: https://github.com/ZPDu/Domain-general-Crowd-Counting-in-Unseen-Scenarios

Postures anomaly tracking and prediction learning model over crowd data analytics.

Innovative technology and improvements in intelligent machinery, transportation facilities, emergency systems, and educational services define the modern era. It is difficult to comprehend the scenario, do crowd analysis, and observe persons. For e-learning-based multiobject tracking and predication framework for crowd data via multilayer perceptron, this article recommends an organized method that takes e-learning crowd-based type data as input, based on usual and abnormal actions and activities. After that, super pixel and fuzzy c mean, for features extraction, we used fused dense optical flow and gradient patches, and for multiobject tracking, we applied a compressive tracking algorithm and Taylor series predictive tracking approach. The next step is to find the mean, variance, speed, and frame occupancy utilized for trajectory extraction. To reduce data complexity and optimization, we applied T-distributed stochastic neighbor embedding (t-SNE). For predicting normal and abnormal action in e-learning-based crowd data, we used multilayer perceptron (MLP) to classify numerous classes. We used the three-crowd activity University of California San Diego, Department of Pediatrics (USCD-Ped), Shanghai tech, and Indian Institute of Technology Bombay (IITB) corridor datasets for experimental estimation based on human and nonhuman-based videos. We achieve a mean accuracy of 87.00%, USCD-Ped, Shanghai tech for 85.75%, and IITB corridor of 88.00% datasets.

Driving pattern analysis to determine driver behaviors for local authority based on cloud using OBD II

Aggressive driving is the main cause of road accidents and it is affected by driving behavior which endanger not only the driver himself but also the people around. It is very significant step to identify such behaviors of the drivers by the local authorities which would help in correcting the behaviors or to understand the root cause of the accidents by analyzing the data recorded by the On Board Diagnostic( OBD ) II device. An aggressive driving behavior is characterized by sudden change inmaneuverings of vehicle which eventually yields non uniform parameters values returned by the ECU (Engine Control Unit) system without any specific reason. In this research work, the real time data is recorded from ECU using OBD II and the accelerometer. The Artificial Intelligenceis used in grouping the different types of data toidentifythe behaviors data on the basis of similarity of datapoints.The purpose of this research work is to identify such drivers and reduce the risk of further accidents.The work identifies the behaviors as bad, normal and aggressive behavior. As the clustering is made on basis crowded data which signifies the similar driving patterns for most of the time in the course of recording, therefore, Density-Based Spatial Clustering of Applications with Noise (DBSCAN) unsupervised learning algorithm was used. The data will be sent to the cloud so that it can be accessed by the authority from any place for further action.ANOVA test is conducted usingIBMSPSS(Statistical Package for the Social Sciences) package to compare and determine the best method to collect data by comparing the means between groups.

Multisize Patched Spatial-Temporal Transformer Network for Short- and Long-Term Crowd Flow Prediction

The prediction of urban crowds is crucial not only to traffic management but also to studies on the city-level social phenomena, such as energy consumption, urban growth, city planning, and epidemic prevention. The challenges of accurately predicting crowd flow come from the non-linear spatial-temporal dependence of crowd flow data, periodic laws, such as daily and weekly periodicity, and external factors, such as weather and holidays. It is even more challenging for most existing short-term prediction models to make an accurate long-term prediction. In this paper, we propose a novel patched Transformer-based sequence-to-sequence model, called MultiSize Patched Spatial-Temporal Transformer Network (MSP-STTN), to incorporate rich and unified context modeling via a self-attention mechanism and global memory learning via a cross-attention mechanism for short- and long-term grid-based crowd flow prediction. In particular, a multisize patched spatial-temporal self-attention Transformer is designed to capture cross-space-time and cross-size contextual dependence of crowd data. The same structured cross-attention Transformer is developed to adaptively learn a global memory for long-term prediction in a responding-to-a-query style without error accumulation. In addition, a categorized space-time expectation is proposed as a unified regional encoding with temporal and external factors and is used as a base prediction for stable training. Furthermore, auxiliary tasks are introduced for promoting feature encoding and leveraging feature consistency to assist in the main prediction task. The experimental results reveal that MSP-STTN is competitive with the state of the art for one-step and multi-step short-term prediction within several hours and achieves practical long-term crowd flow prediction within one day on real-world grid-based crowd data sets TaxiBJ, BikeNYC, and CrowdDensityBJ. Our code and data are available at <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/xieyulai/MSP-STTN</uri> .

Why They Escape: Mining Prioritized Fuzzy Decision Rule in Crowd Evacuation

For safety planning in crowd evacuation, it is important to predict the evacuation decisions made by different individuals and understand the reasons behind these decisions. To this end, this paper proposes an automated approach that can learn prioritized fuzzy decision rules from crowd data to predict and understand the evacuation decisions of a real human. A coevolutionary fuzzy rule miner based on genetic fuzzy-system is designed to select necessary decision features from available ones and learn both rule structure and associated rule parameters from training data. The learned fuzzy rule contains multiple sub-rules, each of which can represent evacuation strategies of different individuals in a given scenario and the features in the fuzzy condition of the sub-rule are organized and evaluated in a sequential order to reflect the priorities of different features. Based on training and testing on four evacuations scenarios of two real-world datasets, it is shown that our proposed approach can learn decision rules that are competitive to the existing evacuation decision models in terms of prediction accuracy. More importantly, it is also demonstrated that our learned rules complying with the proposed prioritized fuzzy rule representation can facilitate the interpretation of evacuation behaviors, such as “herding under zero visibility of exit” and “diminished importance on the distance to exit”, which are aligned to the field observations from real crowd evacuation.

COVID-19 Outbreak Impact on Stock Return in Indonesia and Malaysia

The COVID-19 outbreak drastically impacts the worldwide economy including the stock returns. This study uses Indonesia and Malaysia cases with a total of 141.585 and 219.381 observations in 2020. The measurement of COVID-19 of this study uses daily new confirmed cases and daily death cases from each country. To test the hypothesis, an unbalanced panel and multiple regression are used to test the crowded data. The results show different results for each country. In Indonesia, we found that the COVID-19 outbreak is a strong negative effect on stock return, in contrast, Malaysia found it to be less influence. Furthermore, the effect of Monday and Friday was used to test any impact of the day during the pandemic on stock return. Surprisingly, the effect of Friday affects different from Monday. The weekend effect still occurs in a pandemic situation and the investor mood sentiment to Indonesia’s stock market. This could be one of the aspects for investors consider when making investment decisions during a pandemic.

Multi-level context-driven interaction modeling for human future trajectory prediction

Human trajectory prediction is a challenging task with important applications such as intelligent surveillance and autonomous driving. We recognize that pedestrians in close and distant neighborhoods have different impacts on the person’s decision of future movements. Local scene context and global scene layout also affect the movement decision differently. Existing methods have not adequately addressed these interactions between humans and the multi-level contexts occurring at different spatial and temporal scales. To this end, we propose a multi-level context-driven interaction modeling (MCDIM) method for human future trajectory learning and prediction. Specifically, we construct a multilayer graph attention network (GAT) to model the hierarchical human–human interactions. An extra set of long short-term memory networks is designed to capture the correlations of these human–human interactions at different temporal scales. To model the human–scene interactions, we explicitly extract and encode the global scene layout features and local context features in the neighborhood of the person at each time step and capture the spatial–temporal information of the interactions between human and the local scene contexts. The human–human and human–scene interactions are incorporated into the multi-level GAT-based network for accurate prediction of future trajectories. We have evaluated the method on benchmark datasets: the walking pedestrians dataset provided by ETH Zurich (ETH) and the crowd data provided by the University of Cyprus. The results demonstrate that our MCDIM method outperforms existing methods, being able to generate more accurate and plausible trajectories for pedestrians. The average performance gain is 2 and 3 percentage points in terms of the average displacement error and final displacement error, respectively.

Crowd flow forecasting via agent-based simulations with sequential latent parameter estimation from aggregate observation

Unlike conventional crowd simulations for what-if analysis, agent-based crowd simulations for real-time applications are an emerging research topic and an important tool for better crowd managements in smart cities. Recent studies have attempted to incorporate the real-time crowd observations into crowd simulations for real-time crowd forecasting and management; however, crowd flow forecasting considering individual-level microscopic interactions, especially for large crowds, is still challenging. Here, we present a method that incorporates crowd observation data to forecast a large crowd flow, including thousands of individuals, using a microscopic agent-based model. By sequentially estimating both the crowd state and the latent parameter behind the crowd flows from the aggregate crowd density observation with the particle filter algorithm, the present method estimates and forecasts the large crowd flow using agent-based simulations that incorporate observation data. Numerical experiments, including a realistic evacuation scenario with 5000 individuals, demonstrated that the present method could successfully provide reasonable crowd flow forecasting for different crowd scenarios, even with limited information on crowd movements. These results support the feasibility of real-time crowd flow forecasting and subsequent crowd management, even for large but microscopic crowd problems.

Spatial-Frequency Attention Network for Crowd Counting.

Counting the number of people in crowded scenarios is a crucial task in video surveillance and urban security system. Widely deployed surveillance cameras provide big data for training, a compelling deep learning-based counting network. However, large-scale variations in dense crowds are still not entirely solved. To address this problem, we propose a spatial-frequency attention network (SFANet) for crowd counting in this article. A bottleneck spatial attention module is built to emphasize features in various spatial locations and select a region containing individuals adaptively in the spatial domain. As a complementary, in the frequency domain, a multispectral channel attention module is adopted to obtain a more complete set of frequency components for representing each channel. The two attention modules are combined to focus on the discriminative region and suppress the misleading information by their mutual promotion. Experimental results on five benchmark crowd data sets demonstrate that the SFANet can achieve the state-of-the-art performance in terms of accuracy and robustness.

Real Time Building Evacuation Modeling with an Improved Cellular Automata Method and Corresponding IoT System Implementation

Facility emergence evacuation is often a complicated process under extreme conditions. Most of the buildings today use pre-installed signages to guide the emergence evacuation. However, these guidances are sometimes insufficient or misleading, particularly for evacuating from high-rise buildings or complex buildings, such as schools, hospitals, and stadiums. Following a planned route may lead the crowd to move towards dangers, such as smoke and fire. The future emergency guidance system should be more intelligent and be able to guide people to evacuate with a higher survival possibility. This study proposes a real-time building evacuation model with an improved cellular automata (CA) method. This algorithm combines cellular automata with the potential energy field (PEF) model in fluid dynamic theory (FDT) to choose safe paths for the crowd and reduce the possibility of stampedes. Custom-designed wireless sensors, artificial intelligence (A.I.) enhanced surveillance cameras, intelligent emergency signage systems, and edge computing servers are used to sample fire and crowd data, operate the intelligent evacuation algorithm, and guide the crowd with the signage system in real-time conditions. In addition, we performed the algorithm simulation on a two-dimensional plane generated based on the building structure of the Beijing Capital Airport Hospital. The evacuation drill simulations show that the average escape time is significantly shortened with optimal real-time guidance. In one case, a 72% reduction in evacuation time is achieved compared with evacuation using pre-installed signages. The results also demonstrated that the proposed model and system’s evacuation time reduction performance is particularly good in crowded buildings, such as schools or stadiums.

SIMCD: SIMulated crowd data for anomaly detection and prediction

Smart Crowd management (SCM) solutions can mitigate overcrowding disasters by implementing efficient crowd learning models that can anticipate critical crowd conditions and potential catastrophes. Developing an SCM solution involves monitoring crowds and modelling their dynamics. Crowd monitoring produces vast amounts of data, with features such as densities and speeds, which are essential for training and evaluating crowd learning models. By and large, crowd datasets can be classified as real (e.g., real monitoring of crowds) or synthetic (e.g., simulation of crowds). Using real crowd datasets can produce effective and reliable crowd learning models. However, acquiring real crowd data faces several challenges, including the expensive installation of a sensory infrastructure, the data pre-processing costs and the lack of real datasets that cover particular crowd scenarios. Consequently, crowd management literature has adopted simulation tools for generating synthetic datasets to overcome the challenges associated with their real counterparts. The majority of existing datasets, whether real or synthetic, can be used for crowd counting applications or analysing the activities of individuals rather than collective crowd behaviour. Accordingly, this paper demonstrates the process of generating bespoke synthetic crowd datasets that can be used for crowd anomaly detection and prediction, using the MassMotion crowd simulator. The developed datasets present two types of crowd anomalies; namely, high densities and contra-flow walking direction. These datasets are: SIMulated Crowd Data (SIMCD)-Single Anomaly and SIMCD-Multiple Anomalies for anomaly detection tasks, besides two SIMCD-Prediction datasets for crowd prediction tasks. Furthermore, the paper demonstrates the data preparation (pre-processing) process by aggregating the data and proposing new essential features, such as the level of crowdedness and the crowd severity level, that are useful for developing crowd prediction and anomaly detection models.