This paper presents a novel approach for enhancing video anomaly detection (AD) in crowded scenes using spatio-temporal autoencoders and convolutional LSTM networks. The research aims to address the challenges of detecting and locating anomalies in complex video sequences, particularly in densely populated environments. The proposed method leverages both spatial and temporal dimensions to capture intricate motion patterns and spatial information embedded in video data. The model is designed to efficiently detect and precisely locate anomalies, even in short time frames. The efficacy of the proposed model is evaluated using a benchmark dataset that simulates real-world conditions where millions of video footage need to be continuously monitored in real time. The research highlights the limitations of traditional supervised learning (SL) methods in this context, especially given the massive volume of data and the need for rapid AD. The paper proposes an automated framework for detecting and segmenting pertinent sequences, leveraging deep learning (DL) for efficient feature extraction and representation from video data. The method uses convolutional autoencoders and temporal autoencoders to identify spatial structures and regular temporal patterns, respectively. The ability to automatically fine-tune the model according to the video allows for domain-free and efficient processing. The paper also discusses the challenges of anomaly detection, including the difficulty in defining anomalies and the potential for class imbalance in supervised learning. The proposed approach aims to improve upon earlier labeling methods by reducing labor-intensive efforts and utilizing unsupervised processing. The paper is structured into sections including introduction, related works, proposed methodology, results and discussion, and conclusion and future work. The research contributes to the field of AD by proposing a robust and sophisticated framework that can effectively capture spatial and temporal information within video sequences, enabling the identification and localization of anomalous activities with higher precision in various surveillance and security applications.This paper presents a novel approach for enhancing video anomaly detection (AD) in crowded scenes using spatio-temporal autoencoders and convolutional LSTM networks. The research aims to address the challenges of detecting and locating anomalies in complex video sequences, particularly in densely populated environments. The proposed method leverages both spatial and temporal dimensions to capture intricate motion patterns and spatial information embedded in video data. The model is designed to efficiently detect and precisely locate anomalies, even in short time frames. The efficacy of the proposed model is evaluated using a benchmark dataset that simulates real-world conditions where millions of video footage need to be continuously monitored in real time. The research highlights the limitations of traditional supervised learning (SL) methods in this context, especially given the massive volume of data and the need for rapid AD. The paper proposes an automated framework for detecting and segmenting pertinent sequences, leveraging deep learning (DL) for efficient feature extraction and representation from video data. The method uses convolutional autoencoders and temporal autoencoders to identify spatial structures and regular temporal patterns, respectively. The ability to automatically fine-tune the model according to the video allows for domain-free and efficient processing. The paper also discusses the challenges of anomaly detection, including the difficulty in defining anomalies and the potential for class imbalance in supervised learning. The proposed approach aims to improve upon earlier labeling methods by reducing labor-intensive efforts and utilizing unsupervised processing. The paper is structured into sections including introduction, related works, proposed methodology, results and discussion, and conclusion and future work. The research contributes to the field of AD by proposing a robust and sophisticated framework that can effectively capture spatial and temporal information within video sequences, enabling the identification and localization of anomalous activities with higher precision in various surveillance and security applications.