• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.10
• /
• pp.2643-2657
• /
• 2023

Scene graphs are structured representations that can clearly convey objects and the relationships between them, but are often heavily biased due to the highly skewed, long-tailed relational labeling in the dataset. Indeed, the visual world itself and its descriptions are biased. Therefore, Unbiased Scene Graph Generation (USGG) prefers to train models to eliminate long-tail effects as much as possible, rather than altering the dataset directly. To this end, we propose Geometric and Semantic Improvement (GSI) for USGG to mitigate this issue. First, to fully exploit the feature information in the images, geometric dimension and semantic dimension enhancement modules are designed. The geometric module is designed from the perspective that the position information between neighboring object pairs will affect each other, which can improve the recall rate of the overall relationship in the dataset. The semantic module further processes the embedded word vector, which can enhance the acquisition of semantic information. Then, to improve the recall rate of the tail data, the Class Balanced Seesaw Loss (CBSLoss) is designed for the tail data. The recall rate of the prediction is improved by penalizing the body or tail relations that are judged incorrectly in the dataset. The experimental findings demonstrate that the GSI method performs better than mainstream models in terms of the mean Recall@K (mR@K) metric in three tasks. The long-tailed imbalance in the Visual Genome 150 (VG150) dataset is addressed better using the GSI method than by most of the existing methods.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.18 no.5
• /
• pp.971-980
• /
• 2023

Safety accidents and industrial accidents are constantly occurring in existing industrial sites. In addition, the probability of accidents occurring due to physical and mental fatigue of workers is increasing. Accordingly, it is required to introduce systematic management and various systems for the safety of workers. In this paper, by developing an access control system using bio-metric information at industrial sites, we develop efficient health management and access control management functions for workers. Workers are identified through face recognition for access control, and health status is determined through iris recognition. It aims to improve accuracy and develop a more efficient management system by diagnosing signs of health abnormalities through the congestion of the iris and eyes of workers. Finally, the contents of the development consist of an on-site access control system, an access control program for administrators, and a main server system that diagnoses signs of abnormal health of users.

• International Journal of Computer Science & Network Security
• /
• v.24 no.1
• /
• pp.163-177
• /
• 2024

The fingerprint image quality relies on the clearness of separated ridges by valleys and the uniformity of the separation. The condition of skin still dominate the overall quality of the fingerprint. However, the identification performance of such system is very sensitive to the quality of the captured fingerprint image. Fingerprint image quality analysis and enhancement are useful in improving the performance of fingerprint identification systems. A fuzzy technique is introduced in this paper for both fingerprint image quality analysis and enhancement. First, the quality analysis is performed by extracting four features from a fingerprint image which are the local clarity score (LCS), global clarity score (GCS), ridge_valley thickness ratio (RVTR), and the Global Contrast Factor (GCF). A fuzzy logic technique that uses Mamdani fuzzy rule model is designed. The fuzzy inference system is able to analyse and determinate the fingerprint image type (oily, dry or neutral) based on the extracted feature values and the fuzzy inference rules. The percentages of the test fuzzy inference system for each type is as follow: For dry fingerprint the percentage is 81.33, for oily the percentage is 54.75, and for neutral the percentage is 68.48. Secondly, a fuzzy morphology is applied to enhance the dry and oily fingerprint images. The fuzzy morphology method improves the quality of a fingerprint image, thus improving the performance of the fingerprint identification system significantly. All experimental work which was done for both quality analysis and image enhancement was done using the DB_ITS_2009 database which is a private database collected by the department of electrical engineering, institute of technology Sepuluh Nopember Surabaya, Indonesia. The performance evaluation was done using the Feature Similarity index (FSIM). Where the FSIM is an image quality assessment (IQA) metric, which uses computational models to measure the image quality consistently with subjective evaluations. The new proposed system outperformed the classical system by 900% for the dry fingerprint images and 14% for the oily fingerprint images.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.7
• /
• pp.89-98
• /
• 2024

In this paper, we propose a system for recommending lecture videos at the chapter level, addressing the balance between accuracy and processing speed in chapter-level video recommendations. Specifically, it has been observed that enhancing recommendation accuracy reduces processing speed, while increasing processing speed decreases accuracy. To mitigate this trade-off, a hybrid approach is proposed, utilizing techniques such as TF-IDF, k-means++ clustering, and Graph Neural Networks (GNN). The approach involves pre-constructing clusters based on chapter similarity to reduce computational load during recommendations, thereby improving processing speed, and applying GNN to the graph of clusters as nodes to enhance recommendation accuracy. Experimental results indicate that the use of GNN resulted in an approximate 19.7% increase in recommendation accuracy, as measured by the Mean Reciprocal Rank (MRR) metric, and an approximate 27.7% increase in precision defined by similarities. These findings are expected to contribute to the development of a learning system that recommends more suitable video chapters in response to learners' queries.

• International Journal of Computer Science & Network Security
• /
• v.23 no.12
• /
• pp.187-193
• /
• 2023

Large software systems are being produced with a noticeably higher level of quality with component-based software engineering (CBSE), which places a strong emphasis on breaking down engineered systems into logical or functional components with clearly defined interfaces for inter-component communication. The component-based software engineering is applicable for the commercial products of open-source software. Software metrics play a major role in application development which improves the quantitative measurement of analyzing, scheduling, and reiterating the software module. This methodology will provide an improved result in the process, of better quality and higher usage of software development. The major concern is about the software complexity which is focused on the development and deployment of software. Software metrics will provide an accurate result of software quality, risk, reliability, functionality, and reusability of the component. The proposed metrics are used to assess many aspects of the process, including efficiency, reusability, product interaction, and process complexity. The details description of the various software quality metrics that may be found in the literature on software engineering. In this study, it is explored the advantages and disadvantages of the various software metrics. The topic of component-based software engineering is discussed in this paper along with metrics for software quality, object-oriented metrics, and improved performance.

• KIPS Transactions on Computer and Communication Systems
• /
• v.13 no.1
• /
• pp.31-37
• /
• 2024

Conventional secret sharing techniques often derive shares from randomly generated polynomials or planes, resulting in lengthy and complex shares that are challenging to memorize and/or manage without the aid of a separate computer or specialized device. Modifying existing secret sharing methods to use a predetermined value, such as a memorizable password or bio-metric information, offers a solution. However, this approach raises concerns about security, especially when the predetermined value lacks randomness or has low entropy. In such cases, adversaries may deduce a secret S with just (t - 1) shares by guessing the predetermined value or employing brute force attacks. In this paper, we introduce a share hardening method designed to ensure the security of secret sharing while enabling the use of memorizable passwords or biometric information as predetermined shares.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.1
• /
• pp.41-49
• /
• 2024

Speech recognition technology is continuously advancing and widely used in various fields. In this study, we aimed to investigate the impact of speech data quality on speech recognition models by dividing the dataset into the entire dataset and the top 70% based on Signal-to-Noise Ratio (SNR). Utilizing Seamless M4T and Google Cloud Speech-to-Text, we examined the text transformation results for each model and evaluated them using the Levenshtein Distance. Experimental results revealed that Seamless M4T scored 13.6 in models using data with high SNR, which is lower than the score of 16.6 for the entire dataset. However, Google Cloud Speech-to-Text scored 8.3 on the entire dataset, indicating lower performance than data with high SNR. This suggests that using data with high SNR during the training of a new speech recognition model can have an impact, and Levenshtein Distance can serve as a metric for evaluating speech recognition models.

• Computers and Concrete
• /
• v.33 no.1
• /
• pp.55-75
• /
• 2024

The present study introduces an optimum performance soft computing model for predicting the compressive strength of high-performance concrete (HPC) by comparing models based on conventional (kernel-based, covariance function-based, and tree-based), advanced machine (least square support vector machine-LSSVM and minimax probability machine regressor-MPMR), and deep (artificial neural network-ANN) learning approaches using a common database for the first time. A compressive strength database, having results of 1030 concrete samples, has been compiled from the literature and preprocessed. For the purpose of training, testing, and validation of soft computing models, 803, 101, and 101 data points have been selected arbitrarily from preprocessed data points, i.e., 1005. Thirteen performance metrics, including three new metrics, i.e., a20-index, index of agreement, and index of scatter, have been implemented for each model. The performance comparison reveals that the SVM (kernel-based), ET (tree-based), MPMR (advanced), and ANN (deep) models have achieved higher performance in predicting the compressive strength of HPC. From the overall analysis of performance, accuracy, Taylor plot, accuracy metric, regression error characteristics curve, Anderson-Darling, Wilcoxon, Uncertainty, and reliability, it has been observed that model CS4 based on the ensemble tree has been recognized as an optimum performance model with higher performance, i.e., a correlation coefficient of 0.9352, root mean square error of 5.76 MPa, and mean absolute error of 4.1069 MPa. The present study also reveals that multicollinearity affects the prediction accuracy of Gaussian process regression, decision tree, multilinear regression, and adaptive boosting regressor models, novel research in compressive strength prediction of HPC. The cosine sensitivity analysis reveals that the prediction of compressive strength of HPC is highly affected by cement content, fine aggregate, coarse aggregate, and water content.

• The Journal of the Acoustical Society of Korea
• /
• v.29 no.4
• /
• pp.258-264
• /
• 2010

In this paper, we propose an objective quality feature which is based on the human auditory system to measure vacuum cleaner noise. It is observed that some frequency bands are more sensitive to the human auditory system. Therefore, we divided the audible frequency range of vacuum cleaner noise into a number of frequency bands and the average energy of these bands was calculated. Among a number of average energies, an average energy of a frequency band was selected as the proposed feature. In order to test the performance of the proposed feature, fourteen vacuum cleaners were chosen and the noise was recorded in an anechoic-chamber. Then we performed subjective tests to obtain subjective scores of the noise data using the PCM (paired comparison method) and ACR (absolute category rating) subjective methods. The proposed objective quality feature shows high correlation with the subjective scores.

• Korean Journal of Veterinary Service
• /
• v.47 no.1
• /
• pp.1-7
• /
• 2024

Radiographic left atrial dimension (RLAD) is a valuable metric for assessing left atrial enlargement in dogs. While there have been studies on the use of RLAD and the increase in C-reactive protein (CRP) levels based on heart disease stages, there has been no prior research on the correlation between RLAD and CRP. In this study, the objective was to investigate the relationship between the rise in RLAD as myxomatous mitral valve disease (MMVD) stages advance and the increase in CRP levels with MMVD stage progression. In this study, a total of 30 small-breed dogs were included as subjects. These dogs were diagnosed with MMVD at the American College of Veterinary Internal Medicine (ACVIM) stage B1 or B2, or stage C, based on a comprehensive assessment including physical examination, thoracic radiography, and echocardiography. Measurements of VHS and RLAD were compared to assess any significant differences. There were significant differences in RLAD between dogs with MMVD ACVIM stage B1 and those with stage C. The monocytes and CRP levels showed significant differences between ACVIM stage B1, B2 and ACVIM C. Additionally, a significant correlation was observed between the RLAD and VHS measurements. This underscores the notable association between MMVD stage advancement and elevated monocyte and CRP levels. The RLAD scores exhibited a significant difference among dogs with ACVIM stages B1, B2, and C, and significant variations were also observed in monocyte and CRP levels. These results suggest that monocyte and CRP levels may be a valuable diagnostic indicator for heart disease in dogs during the diagnostic evaluation.