• Title/Summary/Keyword: Behavior detection

Search Result 935, Processing Time 0.031 seconds

A Study on the Usage Pattern Based Detection of Abnormal Behavior in BYOD Environment using Regression (BYOD 환경에서 회귀분석을 활용한 비정상 행위 탐지 방법에 관한 연구)

  • Jurn, Jee-Soo;Kim, Tae-Eun;Jo, Chang-Min;Kim, Hwan-Kuk;Son, Kyung-Ho
    • Annual Conference of KIPS
    • /
    • 2015.10a
    • /
    • pp.646-647
    • /
    • 2015
  • 정보통신의 발달로 개인의 모바일 기기를 업무에 활용하는 BYOD(Bring Your Own Device) 시대가 도래 하면서 기업들은 기밀정보 유출 방지, 접근 제어 및 효율적인 자원 관리를 위해 다양한 시스템을 도입하였다. 하지만 접근제어 정책은 획일화 되어 사용자에게 적용되고 있고, 기기의 잦은 분실과 도난, 낮은 보안성 등으로 인한 보안 위협이 존재하기 때문에 BYOD는 적극적으로 도입되지 못하고 있다. 따라서 개인화된 상황정보 수집을 통하여 유연한 정책 설정 및 비정상 사용자를 탐지 및 통제하는 방법이 필요하다. 본 논문에서는 BYOD 환경에서 발생할 수 있는 비정상 행위를 탐지하기 위해 사용자의 서비스 이용속도 분석하여 비정상 행위를 탐지하는 방안에 대해 논의한다.

Abnormal Human Activity Recognition System Based on CNN For Elderly Home Care (노인 홈 케어를위한 CNN 기반의 비정상 인간 활동 인식 시스템)

  • Valavi, Arezoo;Lee, Hyo Jong
    • Annual Conference of KIPS
    • /
    • 2019.05a
    • /
    • pp.542-544
    • /
    • 2019
  • Changes in a person's health affect one's lifestyle and work activities. According to the World Health Organization (WHO), abnormal activity is growing faster in people aged 60 or more than any other age group in almost every country. This trend steadily continues and expected to increase further in the near future. Abnormal activity put these people at high risk of expected incidents since most of these people live alone. Human abnormal activity analysis is a challenging, useful and interesting problem among the researchers and its particularly crucial task in life and health care areas. In this paper, we discuss the problem of abnormal activities of old people lives alone at home. We propose Convolutional Neural Network (CNN) based model to detect the abnormal behaviors of elderlies by utilizing six simulated action data from daily life actions.

A Study on Detecting Dross in Coating Layer on Hot-dip Galvanizing Steels (용융아연 도금강판의 도금층에 잔류한 드로스 검출에 관한 연구)

  • 김유철;이호종
    • Journal of Surface Science and Engineering
    • /
    • v.36 no.6
    • /
    • pp.466-474
    • /
    • 2003
  • To develop a method of detecting dross in coating layer on hot-dip galvanizing steel, chemical etching behavior of the artificial coating layers with top and bottom dross were investigated. After chemical etching with the mixture of picric acid and sodium thiosulfate, each of the top and bottom dross take its distinct color, and alloy layer in coating is also observed. Defects in the coating layers of HGI(hot rolled galvanized iron), CGI(continuous galvanized steel sheet) and GA(galvannealed steel) were analysed, and methods of dross detection which can be applied to inspection process in manufacture were suggested.

Mechanisms of ASR surface cracking in a massive concrete cylinder

  • Kagimoto, Hiroyuki;Yasuda, Yukihiro;Kawamura, Mitsunori
    • Advances in concrete construction
    • /
    • v.3 no.1
    • /
    • pp.39-54
    • /
    • 2015
  • Relative humidity and strains within a massive concrete cylinder (${\varphi}450mm{\times}900mm$) in the drying and the re-saturating process were measured for elucidating the process of ASR surface cracking in concrete. The expansion behavior of mortars in dry atmospheres with various R.H. values and the resaturating process was revealed. Non- or less-expansive layers were formed in near-surface regions in the concrete cylinder in the drying process, but ASR expansions actively progressed in inner portions. After resaturating, R.H. values of near-surface regions rapidly increased with time, but expansions in the regions were found to be very small. However, in the middle portions, of which R.H. values were kept 80% ~ 90% R.H. in the drying process, expansion actively progressed, resulting in further development of surface cracks in the re-saturating process.

Efficiency of Exponential Deperm Protocol

  • Kim, Yongmin;Kim, Young-Hak;Shin, Kwang-Ho
    • Journal of Magnetics
    • /
    • v.18 no.3
    • /
    • pp.326-330
    • /
    • 2013
  • Magnetic treatment of surface vessels and submarines (Deperm) is required to camouflage them against magnetic detection from enemy marine force. So far, deperm has been accomplished by applying an alternating magnetic field of which amplitude decreases linearly. However, the reduction of the residual flux density in the direction of magnetic field is not linear in the case of the linear protocol, since the ferromagnetic material used to construct a surface vessel, mainly Fe-C, shows a nonlinear behavior in an alternating magnetic field. This is one of main reasons to make an ordinary deperm protocol inefficient. In this paper, we propose the exponential deperming protocol and compare the exponential protocol to conventional linear protocol within the framework of deperm performance. We found out that step number could be reduced in the exponential protocol compare with in the linear protocol, because the larger numbers of deperm steps are dedicated in the irreversible domain process region on the magnetic hysteresis.

Light Scattering Analysis on Coagulation Detection with Magnetic Particles

  • Nahm, Kie B.
    • Current Optics and Photonics
    • /
    • v.2 no.6
    • /
    • pp.623-628
    • /
    • 2018
  • Clotting properties of human blood are important clinical information to monitor for patients with platelet and coagulation disorders. Most devices used to diagnose these disorders utilize blood plasma together with tissue factors and $Ca^{{+}{+}}$ additives. In some instruments, magnetic particles were mixed with blood samples and a rotating magnetic field was applied, resulting in the rotation of magnetic particles, which was probed by impinging light. The working principle seems obvious yet had not been investigated in depth. We modeled the collective behavior of light propagating through magnetic needles, aligned in the direction of the rotating external magnetic field, with scattering light analysis software. Simulation results indicated that the scattering pattern undergoes periodic undulations with respect to the slant angle of the magnetic needles. Also provided is a means of extracting meaningful information from the scattering measurement.

Real-Time Cattle Action Recognition for Estrus Detection

  • Heo, Eui-Ju;Ahn, Sung-Jin;Choi, Kang-Sun
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • v.13 no.4
    • /
    • pp.2148-2161
    • /
    • 2019
  • In this paper, we present a real-time cattle action recognition algorithm to detect the estrus phase of cattle from a live video stream. In order to classify cattle movement, specifically, to detect the mounting action, the most observable sign of the estrus phase, a simple yet effective feature description exploiting motion history images (MHI) is designed. By learning the proposed features using the support vector machine framework, various representative cattle actions, such as mounting, walking, tail wagging, and foot stamping, can be recognized robustly in complex scenes. Thanks to low complexity of the proposed action recognition algorithm, multiple cattle in three enclosures can be monitored simultaneously using a single fisheye camera. Through extensive experiments with real video streams, we confirmed that the proposed algorithm outperforms a conventional human action recognition algorithm by 18% in terms of recognition accuracy even with much smaller dimensional feature description.

A Study on Analysis of Malicious Code Behavior Information for Predicting Security Threats in New Environments

  • Choi, Seul-Ki;Lee, Taejin;Kwak, Jin
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • v.13 no.3
    • /
    • pp.1611-1625
    • /
    • 2019
  • The emergence of new technologies and devices brings a new environment in the field of cyber security. It is not easy to predict possible security threats about new environment every time without special criteria. In other words, most malicious codes often reuse malicious code that has occurred in the past, such as bypassing detection from anti-virus or including additional functions. Therefore, we are predicting the security threats that can arise in a new environment based on the history of repeated malicious code. In this paper, we classify and define not only the internal information obtained from malicious code analysis but also the features that occur during infection and attack. We propose a method to predict and manage security threats in new environment by continuously managing and extending.

Endoscopic Resection of Undifferentiated-type Early Gastric Cancer

  • Lee, Ayoung;Chung, Hyunsoo
    • Journal of Gastric Cancer
    • /
    • v.20 no.4
    • /
    • pp.345-354
    • /
    • 2020
  • Early detection of gastric cancer is crucial because the survival rate can be improved through curative treatment. Although surgery and gastrectomy with lymph node dissection remain as the gold standard for curative treatment, early gastric cancer (EGC) with negligible risk of lymph node metastasis can be treated with endoscopic resection (ER), such as endoscopic submucosal dissection. Among gastric cancers, undifferentiated-type cancer is distinguished from differentiated-type cancer in various aspects in terms of clinical features and pathophysiology. The undifferentiated-type cancer is also known to be associated with an aggressive behavior and a poor prognosis. Therefore, the indication of ER for undifferentiated EGC is limited compared with differentiated-type. Recent studies have reported that ER for undifferentiated EGC is safe and shows favorable short- and long-term outcomes. However, it is necessary to understand the details of the research results and to selectively accept them. In this review, we aimed to evaluate the current practice guidelines and the short-term and long-term outcomes of ER for undifferentiated type EGC.

Effectiveness of e-health systems in improving hypertension management and awareness: a systematic review

  • Alotaibi, Mohamed;Ammad uddin, Mohammad
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • v.16 no.1
    • /
    • pp.173-187
    • /
    • 2022
  • Recent studies have focused on self-management of hypertension using smart devices (cellular phones, tablets, watches). It has proven to be an effective tool for early detection and control of high Blood Pressure (BP) without affecting patients' daily routines. This systematic review surveys the existing self-monitoring systems, evaluate their effectiveness and compares the different approaches. We investigated the current systems in terms of various attributes, including methods used, sample size, type of investigation, inputs/ outputs, rate of success in controlling BP, group of users with higher response rate and beneficiaries, acceptability, and adherence to the system. We identified some limitations, shortcomings, and gaps in the research conducted recently studying the impact of mobile technology on managing hypertension. These shortcomings can generate future research opportunities and enable it to become more realistic and adaptive. We recommended including more observable factors and human behaviors that affect BP. Furthermore, we suggested that vital monitoring/logging and medication tuning are insufficient to improve hypertension control. There is also a need to observe and alter patient behavior and lifestyles.