• International journal of advanced smart convergence
• /
• v.12 no.2
• /
• pp.25-33
• /
• 2023

The purpose of this study is to evaluate the impact of intermediate features on FER performance. To achieve this objective, intermediate features were extracted from the input images at specific layers (FM1~FM4) of the pre-trained network (Resnet-18). These extracted intermediate features and original images were used as inputs to the vision transformer (ViT), and the FER performance was compared. As a result, when using a single image as input, using intermediate features extracted from FM2 yielded the best performance (training accuracy: 94.35%, testing accuracy: 75.51%). When using the original image as input, the training accuracy was 91.32% and the testing accuracy was 74.68%. However, when combining the original image with intermediate features as input, the best FER performance was achieved by combining the original image with FM2, FM3, and FM4 (training accuracy: 97.88%, testing accuracy: 79.21%). These results imply that incorporating intermediate features alongside the original image can lead to superior performance. The findings can be referenced and utilized when designing the preprocessing stages of a deep learning model in FER. By considering the effectiveness of using intermediate features, practitioners can make informed decisions to enhance the performance of FER systems.

• International journal of advanced smart convergence
• /
• v.12 no.2
• /
• pp.56-66
• /
• 2023

The goal of deep learning is to extract complex features from multidimensional data use the features to create models that connect input and output. Deep learning is a process of learning nonlinear features and functions from complex data, and the user data that is employed to train deep learning models has become the focus of privacy concerns. Companies that collect user's sensitive personal information, such as users' images and voices, own this data for indefinite period of times. Users cannot delete their personal information, and they cannot limit the purposes for which the data is used. The study has designed a deep learning method that employs privacy protection technology that uses distributed collaborative learning so that multiple participants can use neural network models collaboratively without sharing the input datasets. To prevent direct leaks of personal information, participants are not shown the training datasets during the model training process, unlike traditional deep learning so that the personal information in the data can be protected. The study used a method that can selectively share subsets via an optimization algorithm that is based on modified distributed stochastic gradient descent, and the result showed that it was possible to learn with improved learning accuracy while protecting personal information.

• International journal of advanced smart convergence
• /
• v.12 no.2
• /
• pp.84-89
• /
• 2023

In recent years, with the breakthrough of Artificial Intelligence (AI) technology in deep learning algorithms such as Generative Adversarial Networks (GANs) and Variational Autoencoders (VAE), AI generation technology has rapidly expanded in various sub-sectors in the art field. 2022 as the explosive year of AI-generated art, especially in the creation of AI-generated art creative design, many excellent works have been born, which has improved the work efficiency of art design. This study analyzed the application design characteristics of AI generation technology in two sub fields of artistic creative design of AI painting and AI animation production , and compares the differences between traditional painting and AI painting in the field of painting. Through the research of this paper, the advantages and problems in the process of AI creative design are summarized. Although AI art designs are affected by technical limitations, there are still flaws in artworks and practical problems such as copyright and income, but it provides a strong technical guarantee in the expansion of subdivisions of artistic innovation and technology integration, and has extremely high research value.

• International journal of advanced smart convergence
• /
• v.12 no.2
• /
• pp.102-107
• /
• 2023

Chinese shadow puppetry is an ancient form of drama with a long history. Known as the "mother of Chinese folk opera", it has rich cultural connotations and artistic value. At the same time, as a classic video game genre, side-scrolling games have many loyal fans around the world. However, in recent years, some previous entertainment cultures such as shadow puppetry are slowly disappearing. In contrast, video games play an increasingly important role in people's entertainment. Combining Chinese traditional culture with video games can be a great way to promote the preservation of these cultures. By making traditional culture more accessible and engaging, developers can help to ensure that these cultures continue to be enjoyed by future generations.The side-scrolling game is a classic game type, and it has many similarities with shadow puppetry. This paper will analyze the similarities and differences between Chinese shadow puppetry and horizontal version games, and try to explore how to organically integrate the two, so as to promote the inheritance and development of traditional culture, and promote cultural innovation and the development of creative industries.

• International journal of advanced smart convergence
• /
• v.11 no.4
• /
• pp.170-176
• /
• 2022

Given the critical importance of enhancing the level of ESG practices, the current research examines the impact of credit card users' pro-environmental characteristics (i.e., environmental self-accountability, eco-label involvement) on their satisfaction with credit card benefits related to green life. That is, this research investigates whether consumers' satisfaction with green credit card benefits varies depending on their environmental self-accountability and eco-label involvement. Specifically, we predict that (1) for consumers with high (vs. low) environmental self-accountability, their satisfaction with credit card benefits related to green life will be higher (hypothesis 1); and (2) when consumers have high (vs. low) eco-label involvement, they will be more likely to be satisfied with credit card benefits related to green life (hypothesis 2). An online survey (N = 293) was conducted to test the two hypotheses. In support of the hypotheses, the results indicate that (1) respondents who had high (vs. low) environmental self-accountability were more satisfied with credit card benefits related to green life, and (2) respondents with high eco-label involvement, as compared to those with low eco-label involvement, reported greater satisfaction with credit card benefits related to green life. We suggest an important insight into how credit card companies approaching ESG issues can increase their consumers' satisfaction with green credit card benefits, considering consumers' individual characteristics such as environmental self-accountability and eco-label involvement.

• International journal of advanced smart convergence
• /
• v.11 no.4
• /
• pp.88-95
• /
• 2022

Vacuum simulation is associated with the prediction and calculation of how materials, pumps and systems will perform using mathematical equations. In this investigation, three different high vacuum systems were simulated and estimated with each vacuum characteristics by VacTran simulator. In each of modelled vacuum systems, selection of gas loads into vessel, combination of rough and high vacuum pumps and dimension of conductance elements were proposed as system variables. In pump station model, the pumping speed to pressures by the combination of root pump was analyzed under the variations of vessel volume. In this study, the effects of outgassing dependent on vessel materials was also simulated and aluminum vessel was estimated to optimum materials. It was obtained from the modelling with diffusion pump that the diameter, length of 50×250[mm]roughing line was characterized as optimum variables to reach the ultimate pressure of 10E-7[torr]. Optimum design factors for vacuum characteristics of modelled vacuum system were achieved by VacTran simulator. Feasibility of VacTran as vacuum simulator was verified and applications of VacTran in high tech process expected to be increased.

• International journal of advanced smart convergence
• /
• v.11 no.4
• /
• pp.20-27
• /
• 2022

In this paper, we propose to improve the designation examination criteria of agencies that provide personal identity proofing based on the resident registration number (RRN), a 13-digit number uniquely assigned by the government to identify Korean citizens. In online commerce, etc., the personal identity proofing agency (PIPA) is a place where online users can prove their personal identity by presenting an alternative means instead of their RRN. The designation examination criteria for PIPAs established in 2012 is a revision of the relevant current laws, and there is a problem in applying the designation examination for alternative means of RRN as the current examination standard. Therefore, in this paper, we propose a method to make the current examination criteria applicable to the newly designated examination of the personal identity proofing service provider based on the current RRN alternative method. According to the current designation examination criteria, only those who satisfy all the examination criteria are designated as the PIPA. However, in reality, it is not in line with the purpose of regulatory reform to require that all examination criteria be satisfied. In the proposed method, it is proposed to apply the standard score system for designation of PIPAs, to make the law current, to secure legal compliance, and to establish a new examination standard to provide a new alternative means of personal identity proofing service. By applying the proposed method to the PIPA designation examination, various alternative means of RRN can be utilized in the online commerce service market.

• International journal of advanced smart convergence
• /
• v.11 no.4
• /
• pp.240-246
• /
• 2022

In order to measure the respiratory rate, one of the major vital signs, many devices have been developed and related studies have been conducted. In particular, as the number of wearers of respirators increases in the COVID-19 pandemic situation, studies have been conducted to measure the respiratory rate of the wearer by attaching an electronic sensor to the respirator, but most of them are cases in which an air flow sensor or a microphone sensor is used. In this study, we design and develop a system that measures the respiratory rate of the wearer using an air pressure sensor in a respirator. Air pressure sensors are inexpensive and consume less power than the other sensors. In addition, since the amount of data required for calculation is small and the algorithm is simple, it is suitable for small-scale and low-power processing devices such as Arduino. We developed an algorithm to measure the respiratory rate of a respirator wearer by analysing air pressure change patterns. In addition, variables that can affect air pressure changes were selected, and experimental scenarios were designed according to the variables. According to the designed scenario, we collected air pressure data while the respirator wearer was breathing. The performance of the developed system was evaluated using the collected data.

• International journal of advanced smart convergence
• /
• v.11 no.4
• /
• pp.57-67
• /
• 2022

The demand for personal training (PT), through which high exercise effects can be achieved within short-term, has recently increased. PT can achieve an exercise amount improvement effect, only if accurate postures are maintained upon performing PT, and exercise with inaccurate postures can cause injuries. However, research is insufficient on exercise amount comparisons and judging exercise accuracy on PT. This study proposes an exercise accuracy measurement algorithm and compares differences in exercise amounts according to exercise postures through experiments using a respiratory gas analyzer. The exercise accuracy measurement algorithm acquires Euler anglesfrom major body parts operated upon exercise through a motion device, based on which the joint angles are calculated. By comparing the calculated joint angles with each reference angle in each exercise step, the status of exercise accuracy is judged. The calculated results of exercise accuracy on squats, lunges, and push-ups showed 0.02% difference in comparison with actually measured results through a goniometer. As a result of the exercise amount comparison experiment according to accurate posture through a respiratory gas analyzer, the exercise amount was higher by 45.19% on average in accurate postures. Through this, it was confirmed that maintaining accurate postures contributes to exercise amount improvement.

• International journal of advanced smart convergence
• /
• v.11 no.4
• /
• pp.247-252
• /
• 2022

The purpose of the domestic physician scientist support program is to promote the development of various biotechnology. Therefore, it can be said that examining whether the purpose of support is being faithfully implemented has an important meaning for the future domestic biotechnology development ecosystem. Therefore, this study limited the subject of analysis to 79 MD-PhD experts who participated or participated in doctor scientist programs at major universities in Korea. Among them, a total of 25 researchers, one researcher from each classroom in parasitology, microbiology, pharmacology, biochemistry, physiology, and anatomy, which had the highest paper citations in the last five years (2016-2021), were selected to examine the relationship between joint research. It was selected as the subject of review. As a result, 25 selected pseudo-scientists(MD-PhD) identified domestic and foreign researchers who participated as co-researchers when publishing in overseas academic journals for the last 5 years(2016-2021), and identified the affiliation and name of the top 5 among them, as well as the pseudo-scientist(MD-PhD), it was possible to identify the relationship of a total of 123 co-researchers(excluding 2 missing values) of the top 5 co-researchers with a high degree of cooperation with respect to the researcher(25 in total), and the collaboration of pseudo-scientists. Relationships, major researchers, and research institutes were examined. Nodexl Basic 2018 ver. (Microsof) was used for the analysis, and the relationship between researchers could be visualized by applying network analysis techniques.