Multi-Target Multi-Camera Tracking and Re-Identification from Detection to Tracking in Real-Time Scenarios

This research project provides a comprehensive literature review in the area of Multi-Target Multi-Camera Tracking (MTMCT) and Re-Identification, with a focus on Real-Time Scenarios. The literature review begins with an exploration of the early developments in MTMCT, outlining the basic concepts and initial challenges. It then delves into key milestones that have shaped the field, focusing on critical aspects such as detection, feature extraction, data association, and tracking, along with the datasets and challenges encountered. The review describes various methods and paradigms used in MTMCT, ranging from traditional tracking approaches to more sophisticated techniques such as single-shot, graph-based approaches, and the integration of edge computing. Special emphasis is given to the adaptation of online and real-time tracking methods, which are crucial in modern applications. In addition, the review covers recent advances in the field. By examining state-of-the-art approaches, the project provides insight into the evolution and current capabilities of MTMCT systems. This comprehensive analysis not only reflects the significant progress made in the field, but also highlights the ongoing challenges and potential directions for future research, particularly in the development of efficient, robust, and ethically responsible MTMCT systems.