MULTICAST: MULTI Confirmation-level Alarm SysTem based on CNN and LSTM to mitigate false alarms for handgun detection in video-surveillance

Abstract

Despite the constant advances in computer vision, integrating modern single-image detectors in real-time handgun alarm systems in video-surveillance is still debatable. Using such detectors still implies a high number of false alarms and false negatives. In this context, most existent studies select one of the latest single-image detectors and train it on a better dataset or use some preprocessing, post-processing or data-fusion approach to further reduce false alarms. However, none of these works tried to exploit the temporal information present in the videos to mitigate false detections. This paper presents a new system, called MULTI Confirmation-level Alarm SysTem based on Convolutional Neural Networks (CNN) and Long Short Term Memory networks (LSTM) (MULTI-CAST), that leverages not only the spacial information but also the temporal information existent in the videos for a more reliable handgun detection. MULTICAST consists of three stages, i) a handgun detection stage, ii) a CNN-based spacial confirmation stage and iii) LSTM-based temporal confirmation stage. The temporal confirmation stage uses the positions of the detected handgun in previous instants to predict its trajectory in the next frame. Our experiments show that MULTI-CAST reduces by 80% the number of false alarms with respect to Faster R-CNN based-single-image detector, which makes it more useful in providing more effective and rapid security responses.

A Survey of Video Surveillance Systems in Smart City

“…Method Reason for False Alarm [126] Deep learning Alarms are recognized as abnormal when unknown or new normality appears [121] Neural network What is considered an anomaly today may not be considered an anomaly tomorrow due to the lack of data [161] Deep learning Post-incident alarm triggering occurs from training crowd safety analysis on only human movement [116,118] In order not to create false alarms in the healthcare system, it is necessary to use a human observer [162] Intruders may deliberately trigger false alarms by covering or tampering with cameras Blockchain smart contracts, multi-factor authentication, and multi-class deep learning can be used to mitigate false alarms. Smart contracts deployed on a blockchain help reduce false alarms by providing a decentralized peer review and tamper-resistant framework for alarm management.…”

Weapon Detection in Surveillance Videos Using Human Inspired Particle Swarm Optimization Algorithm and Support Vector Machine

“…Therefore, intelligent systems are developed by researchers for the automatic detection of risk situations or threats involving firearms [9,10]. The intelligent systems are effective in the situations such as terrorist attacks, gunfire incidents on school grounds, mass shooting, and handgun attacks [11,12]. This article uses HIPSO-SVM model for improving the performance of weapon detection and the major contributions are listed below:…”