• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.2
• /
• pp.702-723
• /
• 2020

The application of machine learning (ML) in intrusion detection has attracted much attention with the rapid growth of information security threat. As an efficient multi-label classifier, kernel extreme learning machine (KELM) has been gradually used in intrusion detection system. However, the performance of KELM heavily relies on the kernel selection. In this paper, a novel multiple kernel extreme learning machine (MKELM) model combining the ReliefF with nature-inspired methods is proposed for intrusion detection. The MKELM is designed to estimate whether the attack is carried out and the ReliefF is used as a preprocessor of MKELM to select appropriate features. In addition, the nature-inspired methods whose fitness functions are defined based on the kernel alignment are employed to build the optimal composite kernel in the MKELM. The KDD99, NSL and Kyoto datasets are used to evaluate the performance of the model. The experimental results indicate that the optimal composite kernel function can be determined by using any heuristic optimization method, including PSO, GA, GWO, BA and DE. Since the filter-based feature selection method is combined with the multiple kernel learning approach independent of the classifier, the proposed model can have a good performance while saving a lot of training time.

• Journal of Korea Multimedia Society
• /
• v.18 no.11
• /
• pp.1268-1280
• /
• 2015

In video surveillance system, the exposure of a person's face is a serious threat to personal privacy. To protect the personal privacy in large amount of videos, an automatic face detection method is required to locate and mask the person's face. However, in real-world surveillance videos, the effectiveness of existing face detection methods could deteriorate due to large variations in facial appearance (e.g., facial pose, illumination etc.) or degraded face (e.g., occluded face, low-resolution face etc.). This paper proposes a new face detection method based on multi-level facial features. In a video frame, different kinds of spatial features are independently extracted, and analyzed, which could complement each other in the aforementioned challenges. Temporal domain analysis is also exploited to consolidate the proposed method. Experimental results show that, compared to competing methods, the proposed method is able to achieve very high recall rates while maintaining acceptable precision rates.

• Journal of Internet Computing and Services
• /
• v.11 no.5
• /
• pp.119-126
• /
• 2010

In this paper, we analyze the robustness of transformed-key asymmetric watermarking system and show its improvement by proposing a new detection method. Based on the assumption that the transformed-key asymmetric watermarking system is under the threat of subtraction attack, we first propose the criterion for the detection performance of the watermarking system and analyze the optimum condition on the system. Next, a new detection method is proposed to improve the detection performance of the system based on the criterion. The proposed improvement makes the system robust to not only subtraction attack but also Wu's attack.

• Proceedings of the KSRS Conference
• /
• v.1
• /
• pp.126-129
• /
• 2006

The marine environment is under considerable threat from intentional or accidental oil spills, ballast water discharged, dredging and infilling for coastal development, and uncontrolled sewage and industrial wastewater discharges. Monitoring spills and illegal oil discharges is an important component in ensuring compliance with marine protection legislation and general protection of the coastal environments. For the monitoring task an image processing system is needed that can efficiently perform the detection and the tracking of oil spills and in this direction a significant amount of research work has taken place mainly with the use of radar (SAR) remote sensing data. In this paper the level set image segmentation technique was tested for the detection of oil spills. Level set allow the evolving curve to change topology (break and merge) and therefore boundaries of particularly intricate shapes can be extracted. Experimental results demonstrated that the level set segmentation can be used for the efficient detection and monitoring of oil spills, since the method coped with abrupt shape’s deformations and splits.

• Journal of Biosystems Engineering
• /
• v.40 no.1
• /
• pp.67-77
• /
• 2015

Purpose: Fungal secondary metabolite (mycotoxin) contamination in foods can pose a serious threat to humans and animals. Spectroscopic techniques have proven to be potential alternative tools for early detection of mycotoxins. Thus, the aim of this review is to provide an overview of the current developments in nondestructive food safety testing techniques, particularly regarding fungal contamination testing in grains, focusing on the application of spectroscopic techniques to this problem. Methods: This review focuses on the use of spectroscopic techniques for the detection of fungi and mycotoxins in agricultural products as reported in the literature. It provides an overview of the characteristics of the main spectroscopic methods and reviews their applications in grain analysis. Results: It was found that spectroscopy has advantages over conventional methods used for fungal contamination detection, particularly when combined with chemometrics. These advantages include the rapidness and nondestructive nature of this approach. Conclusions: While spectroscopy offers many benefits for the detection of mycotoxins in agricultural products, a number of limitations exist, which must be overcome prior to widespread adoption of these techniques.

• Journal of the Korea Society of Computer and Information
• /
• v.20 no.6
• /
• pp.37-42
• /
• 2015

In this paper, we propose malware database in mobile memory for realtime malware URL detection and we support realtime malware URL detection engine, that is control the web service for more secure mobile service. Recently, mobile malware is on the rise and to be new threat on mobile environment. In particular the mobile characteristics, the damage of malware is more important, because it leads to monetary damages for the user. There are many researches in cybercriminals prevention and malware detection, but it is still insufficient. Additionally we propose the method for prevention Smishing within SMS, MMS. In the near future, mobile venders must build the secure mobile environment with fundamental measures based on our research.

• Journal of Information Processing Systems
• /
• v.14 no.1
• /
• pp.6-31
• /
• 2018

With the rapid development of the science and technology, it has been becoming more and more convenient to obtain abundant information via the diverse multimedia medium. However, the contents of the multimedia are easily altered with different editing software, and the authenticity and the integrity of multimedia content are under threat. Forensics technology is developed to solve this problem. We focus on reviewing the blind image forensics technologies for copy-move forgery in this survey. Copy-move forgery is one of the most common manners to manipulate images that usually obscure the objects by flat regions or append the objects within the same image. In this paper, two classical models of copy-move forgery are reviewed, and two frameworks of copy-move forgery detection (CMFD) methods are summarized. Then, massive CMFD methods are mainly divided into two types to retrospect the development process of CMFD technologies, including block-based and keypoint-based. Besides, the performance evaluation criterions and the datasets created for evaluating the performance of CMFD methods are also collected in this review. At last, future research directions and conclusions are given to provide beneficial advice for researchers in this field.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.05a
• /
• pp.198-201
• /
• 2022

Recently, new cyber threats targeting new technologies are increasing, and hackers' attack targets are becoming broader and more intelligent. To counter these attacks, major security companies are using traditional EDR (Endpoint Detection and Response) solutions. However, the conventional method does not consider the context, so there is a limit to the accuracy and efficiency of responding to an advanced attack. In order to improve this problem, the need for a security solution centered on XDR (Extended Detection and Response) has recently emerged. In this study, we present effective threat detection and countermeasures in a changing environment through XDR trends and development roadmaps using machine learning-based context analysis.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.42 no.3
• /
• pp.639-648
• /
• 2017

Electronic warfare systems for extracting information of the threat signals can be employed under the circumstance where the power of the received signal is weak. To precisely and rapidly detect the threat signals, it is required to use methods exploiting whole energy of the received signals instead of conventional methods using a single received signal input. To utilize the whole energy, numerous sizes of windows need to be implemented in a detector for dealing with all possible unknown length of the received signal because it is assumed that there is no preliminary information of the uncooperative signals. However, this grid search method requires too large computational complexity to be practically implemented. In order to resolve this complexity problem, an approach that reduces the number of windows by selecting the smaller number of representative windows can be considered. However, each representative window in this approach needs to cover a certain amount of interval divided from the considering range. Consequently, the discordance between the length of the received signal and the window sizes results in degradation of the detection performance. Therefore, we propose the weighted energy detector which results in improved detection performance comparing with the conventional energy detector under circumstance where the window size is smaller than the length of the received signal. In addition, it is shown that the proposed method exhibits the same performance under other circumstances.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.35 no.4B
• /
• pp.566-575
• /
• 2010

In recent, artificial immune system has become an important research direction in the anomaly detection of networks. The conventional artificial immune systems are usually based on the negative selection that is one of the computational models of self/nonself discrimination. A main problem with self and non-self discrimination is the determination of the frontier between self and non-self. It causes false positive and false negative which are wrong detections. Therefore, additional functions are needed in order to detect potential anomaly while identifying abnormal behavior from analogous symptoms. In this paper, we design novel network attack detection and response schemes based on artificial immune system, and evaluate the performance of the proposed schemes. We firstly generate detector set and design detection and response modules through adopting the interaction between dendritic cells and T-cells. With the sequence of buffer occupancy, a set of detectors is generated by negative selection. The detection module detects the network anomaly with a set of detectors and generates alarm signal to the response module. In order to reduce wrong detections, we also utilize the fuzzy number theory that infers the degree of threat. The degree of threat is calculated by monitoring the number of alarm signals and the intensity of alarm occurrence. The response module sends the control signal to attackers to limit the attack traffic.