• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.20 no.6
• /
• pp.157-165
• /
• 2020

Aiming at the development of neural network and self-driving data set, it is also an idea to improve the performance of network model to detect moving objects by dividing the data set. In Darknet network framework, the YOLOv4 (You Only Look Once v4) network model was used to train and test Udacity data set. According to 7 proportions of the Udacity data set, it was divided into three subsets including training set, validation set and test set. K-means++ algorithm was used to conduct dimensional clustering of object boxes in 7 groups. By adjusting the super parameters of YOLOv4 network for training, Optimal model parameters for 7 groups were obtained respectively. These model parameters were used to detect and compare 7 test sets respectively. The experimental results showed that YOLOv4 can effectively detect the large, medium and small moving objects represented by Truck, Car and Pedestrian in the Udacity data set. When the ratio of training set, validation set and test set is 7:1.5:1.5, the optimal model parameters of the YOLOv4 have highest detection performance. The values show mAP50 reaching 80.89%, mAP75 reaching 47.08%, and the detection speed reaching 10.56 FPS.

• KIPS Transactions on Computer and Communication Systems
• /
• v.11 no.10
• /
• pp.363-372
• /
• 2022

Since the recent COVID-19 Pandemic, the ransomware fandom has intensified along with the expansion of remote work. Currently, anti-virus vaccine companies are trying to respond to ransomware, but traditional file signature-based static analysis can be neutralized in the face of diversification, obfuscation, variants, or the emergence of new ransomware. Various studies are being conducted for such ransomware detection, and detection studies using signature-based static analysis and behavior-based dynamic analysis can be seen as the main research type at present. In this paper, the frequency of ".text Section" Opcode and the Native API used in practice was extracted, and the association between feature information selected using K-means Clustering algorithm, Cosine Similarity, and Pearson correlation coefficient was analyzed. In addition, Through experiments to classify and detect worms among other malware types and Cerber-type ransomware, it was verified that the selected feature information was specialized in detecting specific ransomware (Cerber). As a result of combining the finally selected feature information through the above verification and applying it to machine learning and performing hyper parameter optimization, the detection rate was up to 93.3%.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.30 no.5
• /
• pp.415-425
• /
• 2024

The global maritime industry is developing rapidly owing to the emergence of autonomous ship technology, and interest in utilizing artificial intelligence derived from marine data is increasing. Among the diverse technological developments, ship-route clustering is emerging as an important technology for the commercialization of autonomous ships. Through route clustering, ship-route patterns are extracted from the sea to obtain the fastest and safest route and serve as a basis for the development of a collision-prevention system. High-quality, well-processed data are essential in ensuring the accuracy and efficiency of route-clustering algorithms. In this study, among the various route-clustering methods, we focus on the ship-route-similarity-based clustering method, which can accurately reflect the actual shape and characteristics of a route. To maximize the efficiency of this method, we attempt to formulate an optimal combination of data-preprocessing technologies. Specifically, we combine four methods of measuring similarity between ship routes and three dimensionality-reducing methods. We perform k-means cluster analysis for each combination and then quantitatively evaluate the results using the silhouette index to obtain the best-performing preprocessing combination. This study extends beyond merely identifying the optimal preprocessing technique and emphasizes the importance of extracting meaningful information from a wide range of ocean data. Additionally, this study can be used as a reference for effectively responding to the digital transformation of the maritime and shipping industry in the Fourth Industrial Revolution era.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.46 no.3
• /
• pp.52-58
• /
• 2009

If a moving target has a linear characteristics, the Kalman filter can estimate relatively accurate the location of a target, but this performance depends on how the dynamic status characteristics of the target is accurately modeled. In many practical problems of tracking a maneuvering target, a simple kinematic model can fairly accurately describe the target dynamics for a wide class of maneuvers. However, since the target can exhibit a wide range of dynamic characteristics, no fixed SKF(Simple Kalman filter) can be matched to estimate, to the required accuracy, the states of the target for every specific maneuver. In this paper, a new AKF(Active Kalman filter) is proposed to solve this problem The process noise covariance level of the Kalman filter is adjusted at each time step according to the study result which uses the neural network algorithm. It is demonstrated by means of a computer simulation that the tracking capability of the proposed AKF(Active Kalman filter) is better than that of the SKF(Simple Kalman Filter).

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.37 no.4
• /
• pp.219-231
• /
• 2019

The specifications of signboards are set for each type of signboards, but the shape and size of the signboard actually installed are not uniform. In addition, because the colors of the signboard are not defined, so various colors are applied to the signboard. Methods for recognizing signboards can be thought of as similar methods of recognizing road signs and license plates, but due to the nature of the signboards, there are limitations in that the signboards can not be recognized in a way similar to road signs and license plates. In this study, we proposed a methodology for recognizing plate-type signboards, which are the main targets of illegal and old signboards, and automatically extracting areas of signboards, using the deep learning-based Faster R-CNN algorithm. The process of recognizing flat type signboards through signboard images captured by using smartphone cameras is divided into two sequences. First, the type of signboard was recognized using deep learning to recognize flat type signboards in various types of signboard images, and the result showed an accuracy of about 71%. Next, when the boundary recognition algorithm for the signboards was applied to recognize the boundary area of the flat type signboard, the boundary of flat type signboard was recognized with an accuracy of 85%.

• Journal of Advanced Navigation Technology
• /
• v.11 no.4
• /
• pp.359-370
• /
• 2007

Lane tracking and obstacle avoidance are considered two of the key technologies on an unmanned ground vehicle system. In this paper, we propose a method of lane tracking and obstacle avoidance, which can be expressed as vehicle control, modeling, and sensor experiments. First, obstacle avoidance consists of two parts: a longitudinal control system for acceleration and deceleration and a lateral control system for steering control. Each system is used for unmanned ground vehicle control, which notes the vehicle's location, recognizes obstacles surrounding it, and makes a decision how fast to proceed according to circumstances. During the operation, the control strategy of the vehicle can detect obstacle and perform obstacle avoidance on the road, which involves vehicle velocity. Second, we explain a method of lane tracking by means of a vision system, which consists of two parts: First, vehicle control is included in the road model through lateral and longitudinal control. Second, the image processing method deals with the lane tracking method, the image processing algorithm, and the filtering method. Finally, in this paper, we propose a method for vehicle control, modeling, lane tracking, and obstacle avoidance, which are confirmed through vehicles tests.

• Journal of Digital Contents Society
• /
• v.14 no.4
• /
• pp.419-428
• /
• 2013

Real-time accessibility and agility are required in u-commerce under ubiquitous computing environment. Most of the existing recommendation techniques adopt the method of evaluation based on personal profile, which has been identified with difficulties in accurately analyzing the customers' level of interest and tendencies, as well as the problems of cost, consequently leaving customers unsatisfied. Researches have been conducted to improve the accuracy of information such as the level of interest and tendencies of the customers. However, the problem lies not in the preconstructed database, but in generating new and diverse profiles that are used for the evaluation of the existing data. Also it is difficult to use the unique recommendation method with hierarchy of each customer who has various characteristics in the existing recommendation techniques. Accordingly, this dissertation used the implicit method without onerous question and answer to the users based on the data from purchasing, unlike the other evaluation techniques. We applied FRAT technique which can analyze the tendency of the various personalization and the exact customer.

• Journal of Digital Convergence
• /
• v.16 no.9
• /
• pp.405-412
• /
• 2018

When a film company decides whether to invest or not in a scenario is the appropriate time to predict box office success. In response to market demands, AI based scenario analysis service has been launched, yet the algorithm is by no means perfect. The purpose of this study is to present a prediction model of movie scenario's box office hit based on human brain processing mechanism. In order to derive patterns of visual, auditory, and cognitive stimuli on the time spectrum of box office animation hit, this study applied Weber's law and brain mechanism. The results are as follow. First, the frequency of brain stimulation in the biggest box office movies was 1.79 times greater than that in the failure movies. Second, in the box office success, the cognitive stimuli codes are spread evenly, whereas in the failure, concentrated among few intervals. Third, in the box office success movie, cognitive stimuli which have big cognition load appeared alone, whereas visual and auditory stimuli which have little cognitive load appeared simultaneously.

• Journal of the Institute of Electronics Engineers of Korea TE
• /
• v.39 no.3
• /
• pp.97-104
• /
• 2002

In this paper we describe the method which optimizes the partition of the input space by means of measure of fuzziness for fuzzy neural network. It covers its generation of fuzzy rules for input sub space. It verifies the performance of the system depended on the various time interval of the input. This method divides the input space into several fuzzy regions and assigns a degree of each of the generated rules for the partitioned subspaces from the given data using the Shannon function and fuzzy entropy function generating the optimal knowledge base without the irrelevant rules. In this scheme the basic idea of the fuzzy neural network is to realize the fuzzy rule base and the process of reasoning by neural network and to make the corresponding parameters of the fuzzy control rules be adapted by the steepest descent algorithm. According to the input interval the proposed inference procedure proves that the fast convergence of root mean square error (RMSE) owes to the optimal partition of the input space

• The Journal of the Korea Contents Association
• /
• v.20 no.6
• /
• pp.556-578
• /
• 2020

There are many different researches about how to analyze issues based on real-time news streams. But, there are few researches which analyze issues hierarchically from news articles and even a previous research of hierarchical issue analysis make clustering speed slower as the increment of news articles. In this paper, we propose a hierarchical and incremental clustering for semi real-time issue analysis on news articles. We trained siamese neural network based weighted cosine similarity model, applied this model to k-means algorithm which is used to make word clusters and converted news articles to document vectors by using these word clusters. Finally, we initialized an issue cluster tree from document vectors, updated this tree whenever news articles happen, and analyzed issues in semi real-time. Through the experiment and evaluation, we showed that up to about 0.26 performance has been improved in terms of NMI. Also, in terms of speed of incremental clustering, we also showed about 10 times faster than before.