• Journal of KIBIM
• /
• v.13 no.3
• /
• pp.12-20
• /
• 2023

Deep learning-based anomaly detection technology is used in various fields such as computer vision, speech recognition, and natural language processing. In particular, this technology is applied in various fields such as monitoring manufacturing equipment abnormalities, detecting financial fraud, detecting network hacking, and detecting anomalies in medical images. However, in the field of construction and architecture, research on deep learning-based data anomaly detection technology is difficult due to the lack of digitization of domain knowledge due to late digital conversion, lack of learning data, and difficulties in collecting and processing field data in real time. This study acquires necessary data through IoT (Internet of Things) from the viewpoint of monitoring for environmental management of architectural spaces, converts them into a database, learns deep learning, and then supports anomaly patterns using AI (Artificial Infelligence) deep learning-based anomaly detection. We propose an implementation process. The results of this study suggest an effective environmental anomaly pattern detection solution architecture for environmental management of architectural spaces, proving its feasibility. The proposed method enables quick response through real-time data processing and analysis collected from IoT. In order to confirm the effectiveness of the proposed method, performance analysis is performed through prototype implementation to derive the results.

• Asian Pacific Journal of Cancer Prevention
• /
• v.13 no.11
• /
• pp.5501-5504
• /
• 2012

Cervical cancer is an important woman's health problems worldwide, especially in low socio-economic countries. The aim of this study was to compare the Pap smear screening results between Akha hill tribe and urban women who live in Chiang Rai province, Thailand. Screening was conducted for 1,100 Akha women and 1,100 urban women who came to have the Pap smear at Chiangrai Prachanukroh Hospital and 1 private cytology laboratory from January to June 2008. The demographic characteristics and factors related to abnormal Pap smears of these women were gathered using closed model questionnaires. Abnormal Pap smears were defined according to the Bethesda 2001 system. The results showed that the prevalence of abnormal Pap smears was 12.2% in Akha women and 4.5% in urban women. The highest prevalence of Pap abnormalities was found in the 41-50 years age group in both populations (4.5% in Akha and 1.7% in urban women). In both populations, abnormal Pap smears were found in <21 years age groups. From the questionnaires, the possible risk factors related to the higher prevalence of abnormal Pap smears in Akha women were early age at marriage (${\leq}17$ years), high frequency pregnacies and high parity and no/low education level. In conclusion, cervical cancer control by education and early detection by Pap smear screening is necessary for hill tribe women. More Pap smear screening service units should be set to improve the coverage for the risk group women who got married in young age, especial in ethnic groups.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.36 no.12C
• /
• pp.744-752
• /
• 2011

In intelligent surveillance system, various methods for detecting abnormal behavior were proposed recently. However, most researches are not robust enough to be utilized for actual reality which often has occlusions because of assumption the researches have that individual objects can be tracked. This paper presents a novel method to detect abnormal behavior by analysing major motion of the scene for complex environment in which object tracking cannot work. First, we generate Visual Word and Visual Document from motion information extracted from input video and process them through LDA(Latent Dirichlet Allocation) algorithm which is one of document analysis technique to obtain major motion information(location, magnitude, direction, distribution) of the scene. Using acquired information, we compare similarity between motion appeared in input video and analysed major motion in order to detect motions which does not match to major motions as abnormal behavior.

• Transactions of the Korean Society of Machine Tool Engineers
• /
• v.11 no.2
• /
• pp.103-111
• /
• 2002

Out of all metal-cutting process, the hole-making process is the most widely used. It is estimated to be more than 30% of the total metal-cutting process. It is therefore desirable to monitor and detect drill wear during the hole-drilling process. One important aspect in controlling the drilling process is monitoring drill wear status. There are two systems, Basic system and Online system, to detect the drill wear. Basic system comprised of spindle rotational speed, feed rates, thrust torque and flank wear measured by tool microscope. Outline system comprised of spindle rotational speed feed rates, AE signal, flank wear area measured by computer vision, On-line monitoring system does not need to stop the process to inspect drill wear. Backpropagation neural networks (BPNs) were used for on-line detection of drill wear. The output was the drill wear state which was either usable or failure. This paper deals with an on-line drill wear monitoring system to fit the detection of the abnormal tool state.

• Convergence Security Journal
• /
• v.9 no.2
• /
• pp.41-48
• /
• 2009

Distributed denial of service attack detection for more development and research is underway. The method of using statistical techniques, the normal packets and abnormal packets to identify efficient. In this paper several statistical techniques, using a mix of various offers a way to detect the attack. To verify the effectiveness of the proposed technique, it set packet filtering on router and the proposed DDoS attacks detection method on a Linux router. In result, the proposed technique was detect various attacks and provide normal service mostly.

• Journal of the Korean Society of Industry Convergence
• /
• v.6 no.4
• /
• pp.325-331
• /
• 2003

This paper described the characteristics of the exhaust pressure and proposed the assistant device for detection of misfired cylinder. Misfire, one of abnormal combustion, affects a bad influence of the 3-way catalyst and emits unburned hydrocarbon. Therefore, to prevent these unusual phenomena and eliminate the factor of the environmental pollution, early detection and correction of the misfired cylinder play a very important role. The configuration of assistant device was changed by length and diameter of pipe and analyzed with the install position on the exhaust system. Experimental results showed that the configuration of assistant device is not affected more than length and diameter of pipe and the assistant device is be effective in the detection of misfired cylinder on the gasoline engine.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.8 no.1
• /
• pp.23-31
• /
• 2000

Knock and misfire, kinds of abnormal combustion, are highly undesirable effect on the internal combustion engine. So, it is important to detect these avnormal combuition and control the ignition timing etc. to avoid these mal-effect factors in real engine system. In this study, the system which detects the knock and the misfire using by spark plug is presented. This system is based on the effect of modulation breakdown voltage(BDV) between the spark gaps. The voltage drop between spark plug electrodes, when an electrical breakdown is initiated, depends on the temperature and pressure in combustion chamber. So, we can detect knock and misfire that produce changes in gas temperature and pressure (consequently, its density) using by BDV signal change which carries information about the character of combustion.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.11 no.2
• /
• pp.1-7
• /
• 2003

This paper proposes a method to detect misfired cylinders by the exhaust pressure ascent rate. The misfire is generated by faults of electric system or faults of fuel delivery system. It is one of the abnormal combustions. Therefore, it increases the unburned hydrocarbon and the carbon monoxide and affects a bad influence to the 3-way catalyst. The misfire causes to decrease the power of the engine and increase the consumption of the fuel. Early detection and correction of the misfired cylinders can prevent these unusual phenomena. The misfired cylinders can be detected by the comparison of exhaust pressure ascent rate during each cycle. The exhaust pressure ascent rate is defined as pressure rise per time. Our experimental results showed that the proposed method is effective in the detection of the misfired cylinders on a gasoline engine regardless loads and revolutions of the engine.

• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.35 no.4
• /
• pp.127-135
• /
• 1986

This paper represents a real time algorithm which improves the some drawbacks in the past methods for detection of the QRS conplexes of ECG signals. In the conventional method we can't detect QRS complex and QRS duration more correctly in case of (1) the contaminated ECG with 60Hz noise, muscle noise. (2) the movement of the baseline for a QRS complex. (3) being abnormal QRS complex with prolonging QRS. Therefore, we have proposed a new algorithm which can detect accurate QRS complex detection in case of the contaminated ECG with 60Hz noise, muscle noise, and movement of baseline for QRS complex. Moreover, in case of prolonging QRS we accomplished to detect not only QRS complex but also a single pulse that has a width proportional to QRS duration. This algorithm which is proposed in our paper in our paper in programmed with 6502 assembly language for real time ECG signal processing.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2002.05a
• /
• pp.632-638
• /
• 2002

In novelty detection, one attempts to discriminate abnormal patterns from normal ones. Novelty detection is quite difficult since, unlike usual two class classification problems, only normal patterns are available for training. Auto-Associative Multi-Layer Perceptron (AAMLP) has been shown to provide a good performance based upon the property that novel patterns usually have larger auto-associative errors. In this paper, we give a mathematical analysis of 2-layer AAMLP's output characteristics and empirical results of 2-layer and 4-layer AAMLPs. Various activation functions such as linear, saturated linear and sigmoid are compared. The 2-layer AAMLPs cannot identify non-linear boundaries while the 4-layer ones can. When the data distribution is multi-modal, then an ensemble of AAMLPs, each of which is trained with pre-clustered data is required. This paper contributes to understanding of AAMLP networks and leads to practical recommendations regarding its use.