Abstract. Recognition of human faces from a distance is highly desirable for law-enforcement. This paper evaluates the use of low-cost, high-definition (HD) body worn video cameras for face recognition from a distance. A comparison of HD vs. Standard-definition (SD) video for face recognition from a distance is presented. HD and SD videos of 20 subjects were acquired in different conditions and at varying distances. The evaluation uses three benchmark algorithms: Eigenfaces, Fisherfaces and Wavelet Transforms. The study indicates when gallery and probe images consist of faces captured from a distance, HD video result in better recognition accuracy, compared to SD video. This scenario resembles real-life conditions of video surveillance and law-enforcement activities. However, at a close range, face data obtained from SD video result in similar, if not better recognition accuracy than using HD face data of the same range.
Sensitive data may be stored in different forms. Not only legal owners but also malicious people are interesting of getting sensitive data. Exposing valuable data to others leads to severe Consequences. Customers, organizations, and /or companies lose their money and reputation due to data breaches. There are many reasons for data leakages. Internal threats such as human mistakes and external threats such as DDoS attacks are two main reasons for data loss. In general, data may be categorized based into three kinds: data in use, data at rest, and data in motion. Data Loss Prevention (DLP) are good tools to identify important data. DLP can do analysis for data content and send feedback to administrators to make decision such as filtering, deleting, or encryption. Data Loss Prevention (DLP) tools are not a final solution for data breaches, but they consider good security tools to eliminate malicious activities and protect sensitive information. There are many kinds of DLP techniques, and approximation matching is one of them. Mrsh-v2 is one type of approximation matching. It is implemented and evaluated by using TS dataset and confusion matrix. Finally, Mrsh-v2 has high score of true positive and sensitivity, and it has low score of false negative.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.