• Annual Conference of KIPS
• /
• 2024.05a
• /
• pp.692-695
• /
• 2024

With the recent advancements in artificial intelligence (AI), the performance of deep learning-based audio deepfake technology has significantly improved. This technology has been exploited for criminal activities, leading to various cases of victimization. To prevent such illicit outcomes, this paper proposes a deep learning-based audio deepfake detection model. In this study, we propose CoNSIST, an improved audio deepfake detection model, which incorporates three additional components into the graph-based end-to-end model AASIST: (i) Squeeze and Excitation, (ii) Positional Encoding, and (iii) Reformulated HS-GAL, This incorporation is expected to enable more effective feature extraction, elimination of unnecessary operations, and consideration of more diverse information, thereby improving the performance of the original AASIST. The results of multiple experiments indicate that CoNSIST has enhanced the performance of audio deepfake detection compared to existing models.

• Computers and Concrete
• /
• v.14 no.6
• /
• pp.745-765
• /
• 2014

In this paper, it is aimed to determine the finite element model updating effects on the structural behavior of long span concrete highway bridges. Birecik Highway Bridge located on the 81stkm of Sanliurfa-Gaziantep state highway over Firat River in Turkey is selected as a case study. The bridge consist of fourteen spans, each of span has a nearly 26m. The total bridge length is 380m and width of bridge is 10m. Firstly, the analytical dynamic characteristics such as natural frequencies and mode shapes are attained from finite element analyses using SAP2000 program. After, experimental dynamic characteristics are specified from field investigations using Operational Modal Analysis method. Enhanced Frequency Domain Decomposition method in the frequency domain is used to extract the dynamic characteristics such as natural frequencies, mode shapes and damping ratios. Analytically and experimentally identified dynamic characteristics are compared with each other and finite element model of the bridge is updated to reduce the differences by changing of some uncertain parameters such as section properties, damages, boundary conditions and material properties. At the end of the study, structural performance of the highway bridge is determined under dead load, live load, and dynamic loads before and after model updating to specify the updating effect. Displacements, internal forces and stresses are used as comparison parameters. From the study, it is seen that the ambient vibration measurements are enough to identify the most significant modes of long span highway bridges. Maximum differences between the natural frequencies are reduced averagely from %46.7 to %2.39 by model updating. A good harmony is found between mode shapes after finite element model updating. It is demonstrated that finite element model updating has an important effect on the structural performance of the arch type long span highway bridge. Maximum displacements, shear forces, bending moments and compressive stresses are reduced %28.6, %21.0, %19.22, and %33.3-20.0, respectively.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.18 no.4
• /
• pp.980-997
• /
• 2024

To improve the brightness of images and reveal hidden information in dark areas is the main objective of low-light image enhancement (LLIE). LLIE methods based on deep learning show good performance. However, there are some limitations to these methods, such as the complex network model requires highly configurable environments, and deficient enhancement of edge details leads to blurring of the target content. Single-scale feature extraction results in the insufficient recovery of the hidden content of the enhanced images. This paper proposed an edge detection-based multi-scale feature enhancement network for LLIE (EDMFEN). To reduce the loss of edge details in the enhanced images, an edge extraction module consisting of a Sobel operator is introduced to obtain edge information by computing gradients of images. In addition, a multi-scale feature enhancement module (MSFEM) consisting of multi-scale feature extraction block (MSFEB) and a spatial attention mechanism is proposed to thoroughly recover the hidden content of the enhanced images and obtain richer features. Since the fused features may contain some useless information, the MSFEB is introduced so as to obtain the image features with different perceptual fields. To use the multi-scale features more effectively, a spatial attention mechanism module is used to retain the key features and improve the model performance after fusing multi-scale features. Experimental results on two datasets and five baseline datasets show that EDMFEN has good performance when compared with the stateof-the-art LLIE methods.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.17 no.8
• /
• pp.779-788
• /
• 1992

The Input /output (I /0) subsystem is often the bottleneck in high performance computer system. Generally, system performance evaluation models were enhanced to include the effect of the I/0 system. In this paper, we modeled the terminal servers which are Indispensable devices In distribution of computer resources. We use M /M /1 Queueing model for find out the point of the system performance FIFO buffer sizes In the terminal server arc the Important fanctions of the system design and could be effected to the overall system functions. We have proposed the of optimal buffer sizes in the model of terminal server for increasing the system performance. We analizing the vatting time for terruanl server using Queueing model. and We find out the reference model result from simulation.

• Journal of ILASS-Korea
• /
• v.24 no.3
• /
• pp.105-113
• /
• 2019

In this paper, a study of Urea-SCR System for Dosing Injector for responding to enhanced environmental regulations has been conducted. There is a limit to the experimental approach due to the structural characteristics of the injector. In order to overcome this problem, The analysis was performed assuming unsteady turbulent flow through computational fluid analysis and the internal flow characteristics of the injector were analyzed. By changing the nozzle shape of the injector, the performance factors of the swirl injector by shape were selected and compared. The design parameters were modified by changing the diameter of the nozzle at a constant ratio compared to the base model. Swirl coefficient, outlet mass flow, and sac volume were selected as performance parameters of the injector. The Conv. model to which the taper was applied showed the dominance in mass flow rate, discharge coefficient and swirl because of the smooth fluid flow by shape. Swirl coefficient, outlet mass flow, and sac volume were selected as performance parameters of the injector. As a result of the comparison coefficient derivation with those performance parameters for comparing the performance of the model-specific injector, the Conv-140 model with the nozzle diameter expanded by 140% showed the best value of the comparison coefficient.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2011.05a
• /
• pp.19-19
• /
• 2011

The appraisals of hydrology model behavior for flow and water quality are generally performed through comparison of simulated data with observed ones. To perform appraisal of hydrology model, some criteria are often used, such as coefficient of determination ($R^2$), Nash and Sutcliffe model efficiency coefficient (NSE), index of agreement (d), modified forms of NSE and d, and relative efficiency criteria NSE and d. These criteria are used not only for hydrology model estimations also for various comparisons of two data sets; This NSE has been often used for SWAT calibration. However, it has been known that the NSE value has some limitations in evaluating hydrology at watersheds under monsoon climate because this statistic is largely affected by higher values in the data set. To overcome these limitations, the SWAT auto-calibration module was enhanced with K-means clustering and direct runoff/baseflow modules. However the NSE is still being used in this module to evaluate model performance. Therefore, the SWAT Auto-calibration module was modified to incorporate alternative efficiency criteria into the SWAT K-means/direct runoff-baseflow auto-calibration module. It is expected that this enhanced SWAT auto-calibration module will provide better calibration capability of SWAT model for all flow regime.

• Journal of Korean Society for Quality Management
• /
• v.42 no.4
• /
• pp.617-632
• /
• 2014

Purpose: This study aims to explore the relationship between the focal firm's interdepartmental cooperation and new product development (NPD) performance with focus on the moderating effect of environmental uncertainty. The basic hypothesized model is that there are positively associated relationships. Methods: The proposed research model was tested using structural equation modeling with 601 responses from multi-functional and multiple respondents in Korean manufacturing firms. Multi-group SEM analyses were conducted to explore the degree to which the hypothesized model was equivalent for different levels of environmental uncertainty. Results: Interdepartmental cooperation between R&D and production is positively associated with NPD performance under both higher and lower environmental uncertainties, while one between R&D and marketing is positively associated under only higher environmental uncertainty. Conclusion: This paper determined that NPD performance is positively correlated with R&D-production cooperation in a focal firm, and the relationship between R&D-marketing cooperation and NPD performance is positively moderated by level of environmental uncertainty. Consequently, this study suggests that it is always important for firms to put much effort on R&D-production cooperation for a better NPD performance, while R&D-marketing cooperation should be enhanced especially under higher environmental uncertainty than lower.

• International conference on construction engineering and project management
• /
• 2024.07a
• /
• pp.1256-1263
• /
• 2024

In the realm of Architecture, Engineering, and Construction (AEC) education, various factors play a crucial role in shaping students' acceptance of the learning environments facilitated by visualization technologies, such as virtual reality (VR). Works on leveraging the heterogeneous educational information (i.e., pedagogical data, student performance data, and student survey data) to identify essential factors influencing students' learning experience and performance in virtual environments are still insufficient. This research proposed KnowLearn, an interactive learning assistant system, to integrate an educational knowledge graph (KG) and a locally deployed large language model (LLM) to generate real-time personalized learning recommendations. As the knowledge base of KnowLearn, the educational KG accommodated multi-faceted educational information from twelve perspectives, such as the teaching content, students' academic performance, and their perceived confidence in a specific course from the AEC discipline. A heterogeneous graph attention network (HAN) was utilized to infer the latent information in the KG and, thus, identified the perceived confidence, intention to use, and performance in a relevant quiz as the top three indicators that significantly influenced students' learning outcomes. Based on the information preserved in the KG and learned from the HAN model, the LLM enhanced the personalization of recommendations concerning adopting virtual learning environments while protecting students' privacy. The proposed KnowLearn system is expected to feasibly provide enhanced recommendations on the teaching module design for educators from the AEC domain.

• ETRI Journal
• /
• v.43 no.5
• /
• pp.825-834
• /
• 2021

Based on its simplicity and user-friendly characteristics, OpenMP has become the standard model for programming on shared-memory architectures. Checkpointing-aided parallel execution (CAPE) is an approach that utilizes the discontinuous incremental checkpointing technique (DICKPT) to translate and execute OpenMP programs on distributed-memory architectures automatically. Currently, CAPE implements the OpenMP execution model by utilizing the DICKPT to distribute parallel jobs and their data to slave machines, and then collects the results after executing these distributed jobs. Although this model has been proven to be effective in terms of performance and compatibility with OpenMP on distributed-memory systems, it cannot fully exploit the capabilities of multicore processors. This paper presents a novel execution model for CAPE that utilizes two levels of parallelism. In the proposed model, we add another level of parallelism in the form of multithreaded processes on slave machines with the goal of better exploiting their multicore CPUs. Initial experimental results presented near the end of this paper demonstrate that this model provides significantly enhanced CAPE performance.

• International Commerce and Information Review
• /
• v.12 no.1
• /
• pp.271-295
• /
• 2010

In this study, we investigate the relationships among organizational market orientation, learning orientation, information technology(IT) support for firm resource, IT support for strategy, and balanced scorecard(BSC) performance in export firms. The development of the research model is based on the empirical studies of strategy and resource-based view. The data from the survey was analyzed using Partial Least Squares(PLS). The results from the empirical model suggest that IT support for firm resource is effected by market orientation and learning orientation. And, IT support for strategy is enhanced by IT support for firm resource. Finally, BSC performance of export firms is effected by IT support for strategy.