A Weakly-Supervised Crowd Density Estimation Method Based on Two-Stage Linear Feature Calibration

In a crowd density estimation dataset,the annotation of crowd locations is an extremely laborious task,and they are not taken into the evaluation metrics.In this paper,we aim to reduce the annotation cost of crowd datasets,and propose a crowd density estimation method based on weakly-supervised learning,in the absence of crowd position supervision information,which directly reduces the number of crowds by using the number of pedestrians in the image as the supervised information.For this purpose,we design a new training method,which exploits the correlation between global and local image features by incremental learning to train the network.Specifically,we design a parent-child network(PC-Net)focusing on the global and local image respectively,and propose a linear feature calibration structure to train the PC-Net simultaneously,and the child network learns feature transfer factors and feature bias weights,and uses the transfer factors and bias weights to linearly feature calibrate the features extracted from the Parent network,to improve the convergence of the network by using local features hidden in the crowd images.In addition,we use the pyramid vision transformer as the backbone of the PC-Net to extract crowd features at different levels,and design a global-local feature loss function(L2).We combine it with a crowd counting loss(LC)to enhance the sensitivity of the network to crowd features during the training process,which effectively improves the accuracy of crowd density estimation.The experimental results show that the PC-Net significantly reduces the gap between fullysupervised and weakly-supervised crowd density estimation,and outperforms the comparison methods on five datasets of Shanghai Tech Part A,ShanghaiTech Part B,UCF_CC_50,UCF_QNRF and JHU-CROWD++.

Automated evaluation of parapapillary choroidal microvasculature in crowded optic discs:a controlled,optical coherence tomography angiography study

AIM:To compare superficial and deep vascular properties of optic discs between crowded discs and controls using optical coherence tomography angiography(OCT-A).METHODS:Thirty patients with crowded discs,and 47 control subjects were enrolled in the study.One eye of each individual was included and OCT-A scans of optic discs were obtained in a 4.5×4.5 mm^(2) rectangular area.Radial peripapillary capillary(RPC)density,peripapillary retinal nerve fiber layer(pRNFL)thickness,cup volume,rim area,disc area,cup-to-disc(c/d)area ratio,and vertical c/d ratio were obtained automatically using device software.Automated parapapillary choroidal microvasculature(PPCMv)density was calculated using MATLAB software.When the vertical c/d ratio of the optic disc was absent or small cup,it was considered as a crowded disc.RESULTS:The mean signal strength index of OCT-A images was similar between the crowded discs and control eyes(P=0.740).There was no difference in pRNFL between the two groups(P=0.102).There were no differences in RPC density in whole image(P=0.826)and peripapillary region(P=0.923),but inside disc RPC density was higher in crowded optic discs(P=0.003).The PPCMv density in the inner-hemisuperior region was also lower in crowded discs(P=0.026).The pRNFL thickness was positively correlated with peripapillary RPC density(r=0.498,P<0.001).The inside disc RPC density was negatively correlated with c/d area ratio(r=-0.341,P=0.002).CONCLUSION:The higher inside disc RPC density and lower inner-hemisuperior PPCMv density are found in eyes with crowded optic discs.

Simulation of crowd evacuation under attack considering emotion spreading

In recent years,attacks against crowded places such as campuses and theaters have had a frequent and negative impact on the security and stability of society.In such an event,the crowd will be subjected to high psychological stress and their emotions will rapidly spread to others.This paper establishes the attack-escape evacuation simulation model(AEES-SFM),based on the social force model,to consider emotion spreading under attack.In this model,(1)the attack-escape driving force is considered for the interaction between an attacker and evacuees and(2)emotion spreading among the evacuees is considered to modify the value of the psychological force.To validate the simulation,several experiments were carried out at a university in China.Comparing the simulation and experimental results,it is found that the simulation results are similar to the experimental results when considering emotion spreading.Therefore,the AEES-SFM is proved to be effective.By comparing the results of the evacuation simulation without emotion spreading,the emotion spreading model reduces the evacuation time and the number of casualties by about 30%,which is closer to the real experimental results.The results are still applicable in the case of a 40-person evacuation.This paper provides theoretical support and practical guidance for campus response to violent attacks.

CCM-FL:Covert communication mechanisms for federated learning in crowd sensing IoT

The past decades have witnessed a wide application of federated learning in crowd sensing,to handle the numerous data collected by the sensors and provide the users with precise and customized services.Meanwhile,how to protect the private information of users in federated learning has become an important research topic.Compared with the differential privacy(DP)technique and secure multiparty computation(SMC)strategy,the covert communication mechanism in federated learning is more efficient and energy-saving in training the ma-chine learning models.In this paper,we study the covert communication problem for federated learning in crowd sensing Internet-of-Things networks.Different from the previous works about covert communication in federated learning,most of which are considered in a centralized framework and experimental-based,we firstly proposes a centralized covert communication mechanism for federated learning among n learning agents,the time complexity of which is O(log n),approximating to the optimal solution.Secondly,for the federated learning without parameter server,which is a harder case,we show that solving such a problem is NP-hard and prove the existence of a distributed covert communication mechanism with O(log logΔlog n)times,approximating to the optimal solution.Δis the maximum distance between any pair of learning agents.Theoretical analysis and nu-merical simulations are presented to show the performance of our covert communication mechanisms.We hope that our covert communication work can shed some light on how to protect the privacy of federated learning in crowd sensing from the view of communications.

A Game-Theoretic Approach to Safe Crowd Evacuation in Emergencies

Obstacle removal in crowd evacuation is critical to safety and the evacuation system efficiency. Recently, manyresearchers proposed game theoreticmodels to avoid and remove obstacles for crowd evacuation. Game theoreticalmodels aim to study and analyze the strategic behaviors of individuals within a crowd and their interactionsduring the evacuation. Game theoretical models have some limitations in the context of crowd evacuation. Thesemodels consider a group of individuals as homogeneous objects with the same goals, involve complex mathematicalformulation, and cannot model real-world scenarios such as panic, environmental information, crowds that movedynamically, etc. The proposed work presents a game theoretic model integrating an agent-based model to removethe obstacles from exits. The proposed model considered the parameters named: (1) obstacle size, length, andwidth, (2) removal time, (3) evacuation time, (4) crowd density, (5) obstacle identification, and (6) route selection.The proposed work conducts various experiments considering different conditions, such as obstacle types, obstacleremoval, and several obstacles. Evaluation results show the proposed model’s effectiveness compared with existingliterature in reducing the overall evacuation time, cell selection, and obstacle removal. The study is potentially usefulfor public safety situations such as emergency evacuations during disasters and calamities.

Lightweight Res-Connection Multi-Branch Network for Highly Accurate Crowd Counting and Localization

Crowd counting is a promising hotspot of computer vision involving crowd intelligence analysis,achieving tremendous success recently with the development of deep learning.However,there have been stillmany challenges including crowd multi-scale variations and high network complexity,etc.To tackle these issues,a lightweight Resconnection multi-branch network(LRMBNet)for highly accurate crowd counting and localization is proposed.Specifically,using improved ShuffleNet V2 as the backbone,a lightweight shallow extractor has been designed by employing the channel compression mechanism to reduce enormously the number of network parameters.A light multi-branch structure with different expansion rate convolutions is demonstrated to extract multi-scale features and enlarged receptive fields,where the information transmission and fusion of diverse scale features is enhanced via residual concatenation.In addition,a compound loss function is introduced for training themethod to improve global context information correlation.The proposed method is evaluated on the SHHA,SHHB,UCF-QNRF and UCF_CC_50 public datasets.The accuracy is better than those of many advanced approaches,while the number of parameters is smaller.The experimental results show that the proposed method achieves a good tradeoff between the complexity and accuracy of crowd counting,indicating a lightweight and high-precision method for crowd counting.

Deep Learning Based Efficient Crowd Counting System

Estimation of crowd count is becoming crucial nowadays,as it can help in security surveillance,crowd monitoring,and management for different events.It is challenging to determine the approximate crowd size from an image of the crowd’s density.Therefore in this research study,we proposed a multi-headed convolutional neural network architecture-based model for crowd counting,where we divided our proposed model into two main components:(i)the convolutional neural network,which extracts the feature across the whole image that is given to it as an input,and(ii)the multi-headed layers,which make it easier to evaluate density maps to estimate the number of people in the input image and determine their number in the crowd.We employed the available public benchmark crowd-counting datasets UCF CC 50 and ShanghaiTech parts A and B for model training and testing to validate the model’s performance.To analyze the results,we used two metrics Mean Absolute Error(MAE)and Mean Square Error(MSE),and compared the results of the proposed systems with the state-of-art models of crowd counting.The results show the superiority of the proposed system.

A Survey on Supervised,Unsupervised,and Semi-Supervised Approaches in Crowd Counting

Quantifying the number of individuals in images or videos to estimate crowd density is a challenging yet crucial task with significant implications for fields such as urban planning and public safety.Crowd counting has attracted considerable attention in the field of computer vision,leading to the development of numerous advanced models and methodologies.These approaches vary in terms of supervision techniques,network architectures,and model complexity.Currently,most crowd counting methods rely on fully supervised learning,which has proven to be effective.However,this approach presents challenges in real-world scenarios,where labeled data and ground-truth annotations are often scarce.As a result,there is an increasing need to explore unsupervised and semi-supervised methods to effectively address crowd counting tasks in practical applications.This paper offers a comprehensive review of crowd counting models,with a particular focus on semi-supervised and unsupervised approaches based on their supervision paradigms.We summarize and critically analyze the key methods in these two categories,highlighting their strengths and limitations.Furthermore,we provide a comparative analysis of prominent crowd counting methods using widely adopted benchmark datasets.We believe that this survey will offer valuable insights and guide future advancements in crowd counting technology.

Comparative analysis of manual and programmed annotations for crowd assessment and classification using artificial intelligence

Funding agencies play a pivotal role in bolstering research endeavors by allocating financial resources for data collection and analysis.However,the lack of detailed information regarding the methods employed for data gathering and analysis can obstruct the replication and utilization of the results,ultimately affecting the study’s transparency and integrity.The task of manually annotating extensive datasets demands considerable labor and financial investment,especially when it entails engaging specialized individuals.In our crowd counting study,we employed the web-based annotation tool SuperAnnotate to streamline the human annotation process for a dataset comprising 3,000 images.By integrating automated annotation tools,we realized substantial time efficiencies,as demonstrated by the remarkable achievement of 858,958 annotations.This underscores the significant contribution of such technologies to the efficiency of the annotation process.

Pure and hybrid crowds in crowdfunding markets

Background:Crowdfunding has risen rapidly as a way of raising funds to support projects such as art projects,charity projects,and new ventures.It is very important to understand how crowds in the crowdfunding market are organized to carry out various activities.This study documents and compares two crowd designs for crowdfunding,namely pure crowds,where all crowd members participate as equals,and hybrid crowds,where crowd members are led by an expert investor.The hybrid design is rarely studied in the crowdfunding literature despite its large presence in equity crowdfunding.Methods:We examine industry practices from various countries in terms of crowd designs,review relevant literature on this topic,and develop a conceptual framework for choosing between pure and hybrid crowds.Results:We identify several inefficiencies of pure crowds in crowdfunding platforms and discuss the advantages of hybrid crowds.We then develop a conceptual framework that illustrates the factors for choosing between pure and hybrid crowds.Finally,we discuss the issue of how to manage and regulate lead investors in hybrid crowds.Conclusions:Pure crowds have several shortcomings that could be mitigated by a hybrid crowd design,especially when the proposed project suffers from greater risks,a high degree of information asymmetry,concerns about information leakage,and a high cost of managing the crowds.But for the hybrid crowd to work well,one must carefully design mechanisms for lead investor selection,compensation,and discipline.Our study contributes to the crowdfunding literature and to crowdfunding practice in multiple ways.

Sparse Crowd Flow Analysis of Tawaaf of Kaaba During the COVID-19 Pandemic

The advent of the COVID-19 pandemic has adversely affected the entire world and has put forth high demand for techniques that remotely manage crowd-related tasks.Video surveillance and crowd management using video analysis techniques have significantly impacted today’s research,and numerous applications have been developed in this domain.This research proposed an anomaly detection technique applied to Umrah videos in Kaaba during the COVID-19 pandemic through sparse crowd analysis.Managing theKaaba rituals is crucial since the crowd gathers from around the world and requires proper analysis during these days of the pandemic.The Umrah videos are analyzed,and a system is devised that can track and monitor the crowd flow in Kaaba.The crowd in these videos is sparse due to the pandemic,and we have developed a technique to track the maximum crowd flow and detect any object(person)moving in the direction unlikely of the major flow.We have detected abnormal movement by creating the histograms for the vertical and horizontal flows and applying thresholds to identify the non-majority flow.Our algorithm aims to analyze the crowd through video surveillance and timely detect any abnormal activity tomaintain a smooth crowd flowinKaaba during the pandemic.

Toward Optimal Periodic Crowd Tracking via Unmanned Aerial Vehicle

Crowd management and analysis(CMA)systems have gained a lot of interest in the vulgarization of unmanned aerial vehicles(UAVs)use.Crowd tracking using UAVs is among the most important services provided by a CMA.In this paper,we studied the periodic crowd-tracking(PCT)problem.It consists in usingUAVs to follow-up crowds,during the life-cycle of an open crowded area(OCA).Two criteria were considered for this purpose.The first is related to the CMA initial investment,while the second is to guarantee the quality of service(QoS).The existing works focus on very specified assumptions that are highly committed to CMAs applications context.This study outlined a new binary linear programming(BLP)model to optimally solve the PCT motivated by a real-world application study taking into consideration the high level of abstraction.To closely approach different real-world contexts,we carefully defined and investigated a set of parameters related to the OCA characteristics,behaviors,and theCMAinitial infrastructure investment(e.g.,UAVs,charging stations(CSs)).In order to periodically update theUAVs/crowds andUAVs/CSs assignments,the proposed BLP was integrated into a linear algorithm called PCTs solver.Our main objective was to study the PCT problem fromboth theoretical and numerical viewpoints.To prove the PCTs solver effectiveness,we generated a diversified set of PCTs instances with different scenarios for simulation purposes.The empirical results analysis enabled us to validate the BLPmodel and the PCTs solver,and to point out a set of new challenges for future research directions.

GLCrowd:基于全局-局部注意力的弱监督密集场景人群计数模型

针对人群计数在密集场景下存在背景复杂、尺度变化大等问题,提出了一种结合全局-局部注意力的弱监督密集场景人群计数模型——GLCrowd。首先,设计了一种结合深度卷积的局部注意力模块,通过上下文权重增强局部特征,同时结合特征权重共享获得高频局部信息。其次,利用Vision Transformer(ViT)的自注意力机制捕获低频全局信息。最后,将全局与局部注意力有效融合,并通过回归令牌来完成计数。在Shanghai Tech PartA、Shanghai Tech PartB、UCF-QNRF以及UCF_CC_50数据集上进行了模型测试,MAE分别达到了64.884、8.958、95.523、209.660,MSE分别达到了104.411、16.202、173.453、282.217。结果表明,提出的GLCrowd网络模型在密集场景下的人群计数中具有较好的性能。

基于时延的Flash Crowd控制模型

提出了一种session级别的flash crowd控制策略SGAC(session-granularity admission control),将session控制粒度和request控制粒度相结合,采用请求平均返回时延作为检测和控制的依据.对session采取一旦接受就完成的策略,在实现对服务器过载控制的同时,保护用户session的完整性,并能自动调节新session的准入速率,以提高服务器利用率.采用真实HTTP Log进行模拟,结果表明,SGAC方法能够有效控制服务器过载,保护session的完整性,提高服务器利用率,降低接入端路由器计算开销,保护有价值的交易session.

Knowledge Learning With Crowdsourcing:A Brief Review and Systematic Perspective

Big data have the characteristics of enormous volume,high velocity,diversity,value-sparsity,and uncertainty,which lead the knowledge learning from them full of challenges.With the emergence of crowdsourcing,versatile information can be obtained on-demand so that the wisdom of crowds is easily involved to facilitate the knowledge learning process.During the past thirteen years,researchers in the AI community made great efforts to remove the obstacles in the field of learning from crowds.This concentrated survey paper comprehensively reviews the technical progress in crowdsourcing learning from a systematic perspective that includes three dimensions of data,models,and learning processes.In addition to reviewing existing important work,the paper places a particular emphasis on providing some promising blueprints on each dimension as well as discussing the lessons learned from our past research work,which will light up the way for new researchers and encourage them to pursue new contributions.

Crowdsourced Sampling of a Composite Random Variable: Analysis, Simulation, and Experimental Test

A composite random variable is a product (or sum of products) of statistically distributed quantities. Such a variable can represent the solution to a multi-factor quantitative problem submitted to a large, diverse, independent, anonymous group of non-expert respondents (the “crowd”). The objective of this research is to examine the statistical distribution of solutions from a large crowd to a quantitative problem involving image analysis and object counting. Theoretical analysis by the author, covering a range of conditions and types of factor variables, predicts that composite random variables are distributed log-normally to an excellent approximation. If the factors in a problem are themselves distributed log-normally, then their product is rigorously log-normal. A crowdsourcing experiment devised by the author and implemented with the assistance of a BBC (British Broadcasting Corporation) television show, yielded a sample of approximately 2000 responses consistent with a log-normal distribution. The sample mean was within ~12% of the true count. However, a Monte Carlo simulation (MCS) of the experiment, employing either normal or log-normal random variables as factors to model the processes by which a crowd of 1 million might arrive at their estimates, resulted in a visually perfect log-normal distribution with a mean response within ~5% of the true count. The results of this research suggest that a well-modeled MCS, by simulating a sample of responses from a large, rational, and incentivized crowd, can provide a more accurate solution to a quantitative problem than might be attainable by direct sampling of a smaller crowd or an uninformed crowd, irrespective of size, that guesses randomly.

Abnormal Crowd Behavior Detection Based on the Entropy of Optical Flow

To improve the detection accuracy and robustness of crowd anomaly detection,especially crowd emergency evacuation detection,the abnormal crowd behavior detection method is proposed.This method is based on the improved statistical global optical flow entropy which can better describe the degree of chaos of crowd.First,the optical flow field is extracted from the video sequences and a 2D optical flow histogram is gained.Then,the improved optical flow entropy,combining information theory with statistical physics is calculated from 2D optical flow histograms.Finally,the anomaly can be detected according to the abnormality judgment formula.The experimental results show that the detection accuracy achieved over 95%in three public video datasets,which indicates that the proposed algorithm outperforms other state-of-the-art algorithms.

转发概率递减的改进Crowds系统

Crowds匿名浏览系统中可以在不影响匿名度水平的前提下,通过递减转发概率减小重路由路径长度,提高系统性能。提出利用路径长度期望值递减规律确定转发概率递减比例系数的方法,仿真实验表明,新方案可以在保持原有匿名度的基础上,有效减小重路由路径长度,提高Crowds系统的通信性能。

Crowds系统中新的前驱攻击模型研究

对前驱攻击模型攻击成功的概率和攻击轮数的原理和机制进行了详细的分析,在此基础上,提出了新的统计前驱攻击模型,该模型在每一轮重路由路径建立完毕后,对路径上所有泄密节点的直接前驱进行统计,减少了攻击的重建路径次数.仿真实验表明,与前驱攻击相比,新的统计前驱攻击可以有效地提高攻击的成功概率.

Hybrid tracking model and GSLM based neural network for crowd behavior recognition

Crowd behaviors analysis is the‘state of art’research topic in the field of computer vision which provides applications in video surveillance to crowd safety,event detection,security,etc.Literature presents some of the works related to crowd behavior detection and analysis.In crowd behavior detection,varying density of crowds and motion patterns appears to be complex occlusions for the researchers.This work presents a novel crowd behavior detection system to improve these restrictions.The proposed crowd behavior detection system is developed using hybrid tracking model and integrated features enabled neural network.The object movement and activity in the proposed crowded behavior detection system is assessed using proposed GSLM-based neural network.GSLM based neural network is developed by integrating the gravitational search algorithm with LM algorithm of the neural network to increase the learning process of the network.The performance of the proposed crowd behavior detection system is validated over five different videos and analyzed using accuracy.The experimentation results in the crowd behavior detection with a maximum accuracy of 93%which proves the efficacy of the proposed system in video surveillance with security concerns.