CrowMod: Modeling Crowd Mobility Via Passive Sensing
Crowds involve large groups of people to be gathered in a shared space for a similar communal activity while pursuing distinct individual goals. Crowd modeling is the scientific study and simulation of crowd behavior and movement. The significance of crowd modeling lies in its vast applications, ranging from designing safer public spaces and optimizing event management to enhancing evacuation procedures during emergencies. The applications of crowd modeling are diverse and impactful in areas like urban design, emergency management, event planning, and so on.
Since modeling crowds requires analyzing the behavior of individuals across locations, traditional methods, do not scale. Novel solutions are required that enable scalability while protecting the privacy of the individuals that are being sensed. This thesis focuses on characterizing, detecting, and modeling crowds in public spaces using passive sensing. The core objective is demonstrating how crowds can be modeled using non-intrusive methods such as WiFi sensing. The thesis statement is as follows:
Non-intrusive WiFi sensing can lead to real-time monitoring and modeling of crowds in public spaces while enabling privacy, scalability, and cost-effectiveness.
Specifically, this thesis will leverage anonymous WiFi network traces to extract embedded mobility information at a university campus scale. This includes cleaning and preprocessing WiFi association and dissociation system logs to extract mobility trajectories. This study will also entail crowd modeling and prediction using SOTA ML techniques like LSTMs and unsupervised clustering approaches. Through this, I aim to build a real-time crowd monitoring system for a campus environment that can predict the crowd density given the location and time.
Research Area | Presenter | Title | Keywords |
---|---|---|---|
Computer Science | Berduo, Alan Jesse | Machine learning | |
Engineering | Li, Agnes | Machine learning | |
Computer Science | Shaikh, Aymaan | Machine learning | |
Probability, Statistics, and Machine Learning | Rizvanov, Timur | Machine learning | |
Mathematics and Statistics | Burns, Benjamin | machine learning |