Published Paper Details:

This survey investigates the growing imperative for effective abnormal event detection in video monitoring applications, propelled by the ubiquitous deployment of surveillance cameras in both public and private spaces. Recognizing the challenges posed by labour-intensive human-based surveillance, the study focuses on the current landscape of state-of-theart methodologies in machine learning and deep learning for abnormal event detection. By expanding on prior research, the survey meticulously assesses the techniques' applications within the context of surveillance videos, scrutinizing datasets utilized and providing a nuanced analysis of their strengths and limitations. Through an exhaustive literature review, the paper aims to shed light on the principal challenges inherent in abnormal event detection, distilling key insights to guide future research endeavours in this critical domain. Moreover, the survey contributes to the discourse on bolstering security through intelligent video surveillance systems by synthesizing existing knowledge. In a parallel exploration, the review delves into the realm of crime prediction employing machine learning and deep learning techniques. Examining over 150 articles, the paper elucidates the diverse algorithms applied in predicting crime occurrences, offering insights into patterns and trends. Access to datasets used by researchers is provided, along with an analysis of prevalent approaches and factors influencing criminal activities. The paper identifies potential gaps and proposes future directions to enhance prediction accuracy. As a comprehensive reference for researchers, this overview serves to consolidate the multifaceted landscape of crime prediction using machine learning and deep learning approaches, facilitating advancements in this field.