Collaborative Research: CSR: Small:Accurate and Private Multi-Camera Surveillance System on Time-Sensitive Networks

The rapid growth of smart cities necessitates advanced solutions to improve traffic mobility and public safety. This project proposes a novel multi-camera surveillance system, leveraging a network of distributed smart cameras to capture and analyze streaming video data in real time. By combining the computational power of edge devices and the cloud, this system intelligently processes video streams to address challenges in smart city applications.

The project emphasizes privacy-preserving techniques to ensure sensitive information, such as images of pedestrians and vehicles, is protected while fostering scalable, efficient, and resilient real-time systems. It bridges research domains in systems and networking, machine learning, computer vision, and security and privacy, creating a unified framework for advancing smart city infrastructures.

The project delivers transformative contributions across multiple domains. It introduces innovative unsupervised learning models for tasks such as human and object re-identification and tracking, enabling accurate and efficient analytics in distributed, real-time systems. A novel real-time and resilient cyberinfrastructure is designed with full-stack configurability, addressing system scalability and network performance challenges for large-scale deployments.

Additionally, lightweight cryptographic systems combining advanced cryptographic primitives and Trusted Execution Environments (TEEs) enable privacy-preserving computation for sensitive video data. Beyond technological contributions, the project promotes societal benefits by improving urban services, fostering public trust in privacy-conscious surveillance, and training a new generation of students with skills critical to the systems, networking, data science, and cybersecurity industries.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.