Researchers have examined crowd behaviour in the past by employing a variety of methods including ethnographic studies, computer vision techniques and manual annotation-based data analysis. However, because of the resources to collect, process and analyse data, it remains difficult to obtain large data sets for study. Mobile phones offer easier means for data collection that is easy to analyse and can preserve the user's privacy. The aim of this thesis is to identify and model different qualities of social interactions inside crowds using mobile sensing technology. This Ph.D. research makes three main contributions centred around the mobile sensing and crowd sensing area. Firstly, an open-source licensed mobile sensing framework is developed, named SensingKit, that is capable of collecting mobile sensor data from iOS and Android devices, supporting most sensors available in modern smartphones. The framework has been evaluated in a case study that investigates the pedestrian gait synchronisation phenomenon. Secondly, a novel algorithm based on graph theory is proposed capable of detecting stationary social interactions within crowds. It uses sensor data available in a modern smartphone device, such as the Bluetooth Smart (BLE) sensor, as an indication of user proximity, and accelerometer sensor, as an indication of each user's motion state. Finally, a machine learning model is introduced that uses multi-modal mobile sensor data extracted from Bluetooth Smart, accelerometer and gyroscope sensors. The validation was performed using a relatively large dataset with 24 participants, where they were asked to socialise with each other for 45 minutes. By using supervised machine learning based on gradient-boosted trees, a performance increase of 26.7% was achieved over a proximity-based approach. Such model can be beneficial to the design and implementation of in-the-wild crowd behavioural analysis, design of influence strategies, and algorithms for crowd reconfiguration.

No description available.

The ability to analyse and predict the movement of people in crowded scenarios can be of fundamental importance for tracking across multiple cameras and interaction localisation. In this thesis, we propose a person re-identification method that takes into account the spatial location of cameras using a plan of the locale and the potential paths people can follow in the unobserved areas. These potential paths are generated using two models. In the first, people’s trajectories are constrained to pass through a set of areas of interest (landmarks) in the site. In the second we integrate a goal-driven approach to the Social Force Model (SFM), initially introduced for crowd simulation. SFM models the desire of people to reach specific interest points (goals) in a site, such as exits, shops, seats and meeting points while avoiding walls and barriers. Trajectory propagation creates the possible re-identification candidates, on which association of people across cameras is performed using spatial location of the candidates and appearance features extracted around a person’s head. We validate the proposed method in a challenging scenario from London Gatwick airport and compare it to state-of-the-art person re-identification methods. Moreover, we perform detection and tracking of interacting people in a framework based on SFM that analyses people’s trajectories. The method embeds plausible human behaviours to predict interactions in a crowd by iteratively minimising the error between predictions and measurements. We model people approaching a group and restrict the group formation based on the relative velocity of candidate group members. The detected groups are then tracked by linking their centres of interaction over time using a buffered graph-based tracker. We show how the proposed framework outperforms existing group localisation techniques on three publicly available datasets.

This thesis investigates the use of control theory as a means to study and ultimately control the psychological dynamics of people in a crowd. Gustav LeBon's suggestibility theory, a well-known account of collective behaviour, is used to develop a discrete-time nonlinear model of psychological crowd behavior that, consistent with suggestibility theory, is open-loop unstable. As a first attempt to stabilize the dynamics, linear observer-based output-feedback techniques and a collection of simple nonlinear control strategies are pursued. The poor performance afforded by these schemes motivates an agent-oriented control strategy in which authoritative figures, termed control agents, are interspersed within the crowd and, similar to the technique of feedback linearization, use knowledge of the system dynamics to issue signals that propagate through the crowd to drive specific components of the state to zero. It is shown that if these states are chosen judiciously then it follows that a collection of other state signals are, themselves, zero. This realization is used to develop a stability result for a simple crowd structure and this result is, in turn, used as a template to develop similar results for crowds of greater complexity. Simulations are used to verify the functionality of the reported schemes and the advantages of using multiple control agents, instead of a single control agent, are emphasized. While the mathematical study of complex social phenomena, including crowds, is prefixed by an assortment of unique challenges, the main conclusion of this thesis is that control theory is a potentially powerful framework to study the underlying dynamics at play in such systems.

This thesis investigates the use of control theory as a means to study and ultimately control the psychological dynamics of people in a crowd. Gustav LeBon's suggestibility theory, a well-known account of collective behaviour, is used to develop a discrete-time nonlinear model of psychological crowd behavior that, consistent with suggestibility theory, is open-loop unstable. As a first attempt to stabilize the dynamics, linear observer-based output-feedback techniques and a collection of simple nonlinear control strategies are pursued. The poor performance afforded by these schemes motivates an agent-oriented control strategy in which authoritative figures, termed control agents, are interspersed within the crowd and, similar to the technique of feedback linearization, use knowledge of the system dynamics to issue signals that propagate through the crowd to drive specific components of the state to zero. It is shown that if these states are chosen judiciously then it follows that a collection of other state signals are, themselves, zero. This realization is used to develop a stability result for a simple crowd structure and this result is, in turn, used as a template to develop similar results for crowds of greater complexity. Simulations are used to verify the functionality of the reported schemes and the advantages of using multiple control agents, instead of a single control agent, are emphasized. While the mathematical study of complex social phenomena, including crowds, is prefixed by an assortment of unique challenges, the main conclusion of this thesis is that control theory is a potentially powerful framework to study the underlying dynamics at play in such systems.

No / The first recorded crowdsourcing activity was in 1714 [1], with intermittent public event occurrences up until the millennium when such activities become widespread, spanning multiple domains. Crowdsourcing, however, is relatively novel as a methodology within virtual environment studies, in archaeology, and within the heritage domains where this research is focused. The studies that are being conducted are few and far between in comparison to other areas. This paper aims to develop a recent concept in crowdsourcing work termed `crowd behaviour mining' [2] using virtual environments, and to develop a unique concept in crowdsourcing activities that can be applied beyond the case studies presented here and to other domains that involve human behaviour as independent variables. The case studies described here use data from experiments involving separate heritage projects and conducted during two Royal Society Summer Science Exhibitions, in 2012 and 2015 respectively. `Crowd Behaviour Mining' analysis demonstrated a capacity to inform research in respect of potential patterns and trends across space and time as well as preferences between demographic user groups and the influence of experimenters during the experiments.