Search Results (6)

Shoplifting poses a significant challenge for shop owners as well as other stakeholders, including law enforcement agencies. In recent years, the task of shoplifting detection has gained the interest of researchers due to video surveillance generating vast quantities of data that cannot be processed in real-time by human staff. In previous studies, different datasets and methods have been developed for the task of shoplifting detection. However, there is a lack of a large benchmark dataset containing different behaviors of shoplifting and standard methods for the task of shoplifting detection. To overcome this limitation, in this study, a large benchmark dataset has been developed, having 900 instances with 450 cases of shoplifting and 450 of non-shoplifting with manual annotation based on five different ways of shoplifting. Moreover, a method for the detection of shoplifting is proposed for evaluating the developed dataset. The dataset is also evaluated with methods as baseline methods, including 2D CNN and 3D CNN. Our proposed method, which is a combination of Inception V3 and BILSTM, outperforms all baseline methods with 81 % accuracy. The developed dataset will be publicly available to foster in various areas related to human activity recognition. These areas encompass the development of systems for detecting behaviors such as robbery, identifying human movements, enhancing safety measures, and detecting instances of theft. Full article

The rise in crime rates in many parts of the world, coupled with advancements in computer vision, has increased the need for automated crime detection services. To address this issue, we propose a new approach for detecting suspicious behavior as a means of preventing shoplifting. Existing methods are based on the use of convolutional neural networks that rely on extracting spatial features from pixel values. In contrast, our proposed method employs object detection based on YOLOv5 with Deep Sort to track people through a video, using the resulting bounding box coordinates as temporal features. The extracted temporal features are then modeled as a time-series classification problem. The proposed method was tested on the popular UCF Crime dataset, and benchmarked against the current state-of-the-art robust temporal feature magnitude (RTFM) method, which relies on the Inflated 3D ConvNet (I3D) preprocessing method. Our results demonstrate an impressive 8.45-fold increase in detection inference speed compared to the state-of-the-art RTFM, along with an F1 score of 92%,outperforming RTFM by 3%. Furthermore, our method achieved these results without requiring expensive data augmentation or image feature extraction. Full article

Shoplifting is a major problem for shop owners and many other parties, including the police. Video surveillance generates huge amounts of information that staff cannot process in real time. In this article, the problem of detecting shoplifting in video records was solved using a classifier, which was a hybrid neural network. The hybrid neural network included convolutional and recurrent ones. The convolutional network was used to extract features from the video frames. The recurrent network processed the time sequence of the video frames features and classified the video fragments. In this work, gated recurrent units were selected as the recurrent network. The well-known UCF-Crime dataset was used to form the training and test datasets. The classification results showed a high accuracy of 93%, which was higher than the accuracy of the classifiers considered in the review. Further research will focus on the practical implementation of the proposed hybrid neural network. Full article

Retail shoplifting is one of the most prevalent forms of theft and has accounted for over one billion GBP in losses for UK retailers in 2018. An automated approach to detecting behaviours associated with shoplifting using surveillance footage could help reduce these losses. Until recently, most state-of-the-art vision-based approaches to this problem have relied heavily on the use of black box deep learning models. While these models have been shown to achieve very high accuracy, this lack of understanding on how decisions are made raises concerns about potential bias in the models. This limits the ability of retailers to implement these solutions, as several high-profile legal cases have recently ruled that evidence taken from these black box methods is inadmissible in court. There is an urgent need to develop models which can achieve high accuracy while providing the necessary transparency. One way to alleviate this problem is through the use of social signal processing to add a layer of understanding in the development of transparent models for this task. To this end, we present a social signal processing model for the problem of shoplifting prediction which has been trained and validated using a novel dataset of manually annotated shoplifting videos. The resulting model provides a high degree of understanding and achieves accuracy comparable with current state of the art black box methods. Full article

Experimental studies support the conventional belief that people behave more aggressively whilst under the influence of alcohol. To examine how these experimental findings manifest in real life situations, this study uses a method for estimating evidence for causality with observational data—‘situational decomposition’ to examine the association between alcohol consumption and crime in young adults from the Avon Longitudinal Study of Parents and Children. Self-report questionnaires were completed at age 24 years to assess typical alcohol consumption and frequency, participation in fighting, shoplifting and vandalism in the previous year, and whether these crimes were committed under the influence of alcohol. Situational decomposition compares the strength of two associations, (1) the total association between alcohol consumption and crime (sober or intoxicated) versus (2) the association between alcohol consumption and crime committed while sober. There was an association between typical alcohol consumption and total crime for fighting [OR (95% CI): 1.47 (1.29, 1.67)], shoplifting [OR (95% CI): 1.25 (1.12, 1.40)], and vandalism [OR (95% CI): 1.33 (1.12, 1.57)]. The associations for both fighting and shoplifting had a small causal component (with the association for sober crime slightly smaller than the association for total crime). However, the association for vandalism had a larger causal component. Full article

Crime generates significant losses, both human and economic. Every year, billions of dollars are lost due to attacks, crimes, and scams. Surveillance video camera networks generate vast amounts of data, and the surveillance staff cannot process all the information in real-time. Human sight has critical limitations. Among those limitations, visual focus is one of the most critical when dealing with surveillance. For example, in a surveillance room, a crime can occur in a different screen segment or on a distinct monitor, and the surveillance staff may overlook it. Our proposal focuses on shoplifting crimes by analyzing situations that an average person will consider as typical conditions, but may eventually lead to a crime. While other approaches identify the crime itself, we instead model suspicious behavior—the one that may occur before the build-up phase of a crime—by detecting precise segments of a video with a high probability of containing a shoplifting crime. By doing so, we provide the staff with more opportunities to act and prevent crime. We implemented a 3DCNN model as a video feature extractor and tested its performance on a dataset composed of daily action and shoplifting samples. The results are encouraging as the model correctly classifies suspicious behavior in most of the scenarios where it was tested. For example, when classifying suspicious behavior, the best model generated in this work obtains precision and recall values of 0.8571 and 1 in one of the test scenarios, respectively. Full article