• Journal of KIISE:Software and Applications
• /
• v.32 no.3
• /
• pp.141-152
• /
• 2005

A new emotion-based image retrieval method is proposed in this paper. Query emotion descriptors called query color code and query gray code are designed based on the human evaluation on 13 emotions('like', 'beautiful', 'natural', 'dynamic', 'warm', 'gay', 'cheerful', 'unstable', 'light' 'strong', 'gaudy' 'hard', 'heavy') when 30 random patterns with different color, intensity, and dot sizes are presented. For emotion image retrieval, once a query emotion is selected, associated query color code and query gray code are selected. Next, DB color code and DB gray code that capture color and, intensify and dot size are extracted in each database image and a matching process between two color codes and between two gray codes are peformed to retrieve relevant emotion images. Also, a new relevance feedback method is proposed. The method incorporates human intention in the retrieval process by dynamically updating weights of the query and DB color codes and weights of an intra query color code. For the experiments over 450 images, the number of positive images was higher than that of negative images at the initial query and increased according to the relevance feedback.

• Journal of Biomedical Engineering Research
• /
• v.37 no.2
• /
• pp.61-67
• /
• 2016

This paper presented a method for random forest based the arrhythmia classification using both heart rate (HR) and heart rate variability (HRV) features. We analyzed the MIT-BIH arrhythmia database which contains half-hour ECG recorded from 48 subjects. This study included not only the linear features but also non-linear features for the improvement of classification performance. We classified abnormal ECG using mean_NN (mean of heart rate), SD1/SD2 (geometrical feature of poincare HRV plot), SE (spectral entropy), pNN100 (percentage of a heart rate longer than 100 ms) affecting accurate classification among combined of linear and nonlinear features. We compared our proposed method with Neural Networks to evaluate the accuracy of the algorithm. When we used the features extracted from the HRV as an input variable for classifier, random forest used only the most contributed variable for classification unlike the neural networks. The characteristics of random forest enable the dimensionality reduction of the input variables, increase a efficiency of classifier and can be obtained faster, 11.1% higher accuracy than the neural networks.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.9 no.1
• /
• pp.67-74
• /
• 2016

In this paper, we propose a watermarking technology for a mixed music. The mixed music means recreated music that contained a number of musics in one audio clip. Royalty associated with the audio content is typically imposed by the full audio content. However, the calculation of royalties gives rise to conflict between copyright holders and users in the mixed music because it uses not full audio content but a fraction of that. To solve the conflict related with the mixed music, we propose a audio watermarking technique that inserts different watermarks for each audio in the audio that make up the mixed music. The proposed watermarking scheme might have poor SNR (signal to noise ratio) to embed to each audio clip. To overcome poor SNR problem, we used inaudible pseudo random sequence which modifies typical pseudo random sequence to canonical signed digit (CSD) form. The proposed method verifies the performance by each watermark extraction and the time internal estimation valies from the mixed music.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.29 no.6
• /
• pp.1235-1242
• /
• 2019

While several machine learning technique has been implemented for Android malware categorization, there is still difficulty in analyzing due to overfitting problem and including of un-executable code, etc. In this paper, we introduce our implemented tool to address these problems. Tool is consists of approximately 1,500 lines of Java code, and perform Flow analysis on set of APIs, or on control flow graph. Our tool groups all the API by its relationship and only perform analysis on actually executing code. Using our tool, we grouped 39032 APIs into 4972 groups, and 12123 groups with result of including class names. We collected 7,000 APKs from 7 families and evaluated our feature reduction technique, and we also reduced features again with selecting APIs that have frequency more than 20%. We finally reduced features to 263-numbers of feature for our collected APKs.

• The KIPS Transactions:PartB
• /
• v.14B no.6
• /
• pp.461-472
• /
• 2007

The emphasis in genome projects has Been moving towards the sequence analysis in order to extract biological "meaning"(e.g., evolutionary history of particular molecules or their functions) from the sequence. Especially. palindromic or direct repeats that appear in a sequence have a biophysical meaning and the problem is to recognize interesting patterns and configurations of words(strings of characters) over complementary alphabets. In this paper, we propose an algorithm to identify variable length palindromic pairs(longer than a threshold), where we can allow gaps(distance between words). The algorithm is called palindrome algorithm(PA) and has O(N) time complexity. A palindromic pair consists of a hairpin structure. By composing collected palindromic pairs we build n-pair palindromic patterns. In addition, we dot some of the longest pairs in a circle to represent the structure of a DNA sequence. We run the algorithm over several selected genomes and the results of E.coli K12 are presented. There existed very long palindromic pair patterns in the genomes, which hardly occur in a random sequence.

• The Journal of Korean Association of Computer Education
• /
• v.22 no.3
• /
• pp.55-65
• /
• 2019

In programming language learning system of flip learning methods, the evaluation of cyber lectures generally proceeds from online to multiple choice questions. In this case, the questions are randomly extracted from the question bank and given to individual learners. In order for these evaluation results to be reflected in the grades, the equity of the examination question is more important than anything else. Especially in the programming language subject, the degree of difficulty that learners think can be different depending on the type of problem. In this paper, we classify the types of multiple-choice problems into two categories, and manage the difficulty level by each type. And we propose a question selection algorithm that considers both difficulty level and type of question. Considering the characteristics of the programming language, experimental results show that the proposed algorithm is more efficient and fair than the conventional method.

• The Journal of Bigdata
• /
• v.3 no.2
• /
• pp.11-18
• /
• 2018

This study is a study on domain automatic classification for domain - based quality diagnosis which is a key element of big data quality diagnosis. With the increase of the value and utilization of Big Data and the rise of the Fourth Industrial Revolution, the world is making efforts to create new value by utilizing big data in various fields converged with IT such as law, medical, and finance. However, analysis based on low-reliability data results in critical problems in both the process and the result, and it is also difficult to believe that judgments based on the analysis results. Although the need of highly reliable data has also increased, research on the quality of data and its results have been insufficient. The purpose of this study is to shorten the work time to automizing the domain classification work which was performed from manually to using machine learning in the domain - based quality diagnosis, which is a key element of diagnostic evaluation for improving data quality. Extracts information about the characteristics of the data that is stored in the database and identifies the domain, and then featurize it, and automizes the domain classification using machine learning. We will use it for big data quality diagnosis and contribute to quality improvement.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.20 no.2
• /
• pp.145-156
• /
• 2010

Now a days, the development of TV and internet like communicational technique makes IPTV service which combines internet with multimedia contents increase. But when a user gets service in specific place, the certification process and user's ID check in IPTV service is complicate so that there occurs communicational difficulty like increasing illegal users and service delay etc. This paper proposes communication security mechanism to prevent Clone attack which happens in wireless section by efficiently extracting illegal user. The proposed mechanism performs key distribution procedure, inter certification procedure, and key initiation procedure by putting security agent in RFID-USB for RFID tags users use to perform plug-and-plug function. Also, the proposed mechanism updates the hased token value by its ID and the random number which RFID-USB creates whenever a user accesses in the area of RFID-USB so that it protects reply attack and man-in-the-middle attack which happen often in the area of wireless section.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.12
• /
• pp.1528-1534
• /
• 2019

Recently, due to the popularization of 3D depth cameras, new researches and opportunities have been made in research conducted on RGB images, but estimation of human hand pose is still classified as one of the difficult topics. In this paper, we propose a robust estimation method of human hand pose from various input 3D depth images using a learning algorithm. The proposed approach first generates a skeleton-based hand model and then aligns the generated hand model with three-dimensional point cloud data. Then, using a random forest-based learning algorithm, the hand pose is strongly estimated from the aligned hand model. Experimental results in this paper show that the proposed hierarchical approach makes robust and fast estimation of human hand posture from input depth images captured in various indoor and outdoor environments.

• Journal of the Korea Convergence Society
• /
• v.13 no.2
• /
• pp.249-255
• /
• 2022

This study proposed an analysis framework for real-time prediction of CNC processing defects using machine learning-based models that are recently attracting attention as processing defect prediction methods, and applied it to CNC machines. Analysis shows that the XGBoost, CatBoost, and LightGBM models have the same best accuracy, precision, recall, F1 score, and AUC, of which the LightGBM model took the shortest execution time. This short run time has practical advantages such as reducing actual system deployment costs, reducing the probability of CNC machine damage due to rapid prediction of defects, and increasing overall CNC machine utilization, confirming that the LightGBM model is the most effective machine learning model for CNC machines with only basic sensors installed. In addition, it was confirmed that classification performance was maximized when an ensemble model consisting of LightGBM, ExtraTrees, k-Nearest Neighbors, and logistic regression models was applied in situations where there are no restrictions on execution time and computing power.