• Journal of the Korean Society of Propulsion Engineers
• /
• v.23 no.2
• /
• pp.38-45
• /
• 2019

The present work suggests a numerical approach using a thermoacoustic network model for the eigenvalue calculation of thermoacoustic instability problems in an aero gas turbine combustor. The model is developed based on the conservation laws for mass, momentum, and energy between acoustic network elements with an area change. Acoustic field in a practical aero gas turbine combustor which has a complicated flow path is analyzed using the current model. The predictive capabilities of the current modeling approach are compared with the acoustic characteristics calculated using Helmholtz solver based on 3D finite element method(FEM).

• The Journal of Korea Robotics Society
• /
• v.13 no.1
• /
• pp.8-15
• /
• 2018

This paper proposes a convolutional neural network model for distinguishing areas occupied by obstacles from a LiDAR image converted from a 3D point cloud. The channels of a LiDAR image used as input consist of the distances to 3D points, the reflectivities of 3D points, and the heights of 3D points from the ground. The proposed model uses a LiDAR image as an input and outputs a result of a segmented LiDAR image. The proposed model adopts refinement modules with skip connections to segment a LiDAR image. The refinement modules with skip connections in the proposed model make it possible to construct a complex structure with a small number of parameters than a convolutional neural network model with a linear structure. Using the proposed model, it is possible to distinguish areas in a LiDAR image occupied by obstacles such as vehicles, pedestrians, and bicyclists. The proposed model can be applied to recognize surrounding obstacles and to search for safe paths.

• International Journal of Computer Science & Network Security
• /
• v.23 no.9
• /
• pp.47-54
• /
• 2023

For Bengali music emotion classification, deep learning models, particularly CNN and RNN are frequently used. But previous researches had the flaws of low accuracy and overfitting problem. In this research, attention-based Conv1D and BiGRU model is designed for music emotion classification and comparative experimentation shows that the proposed model is classifying emotions more accurate. We have proposed a Conv1D and Bi-GRU with the attention-based model for emotion classification of our Bengali music dataset. The model integrates attention-based. Wav preprocessing makes use of MFCCs. To reduce the dimensionality of the feature space, contextual features were extracted from two Conv1D layers. In order to solve the overfitting problems, dropouts are utilized. Two bidirectional GRUs networks are used to update previous and future emotion representation of the output from the Conv1D layers. Two BiGRU layers are conntected to an attention mechanism to give various MFCC feature vectors more attention. Moreover, the attention mechanism has increased the accuracy of the proposed classification model. The vector is finally classified into four emotion classes: Angry, Happy, Relax, Sad; using a dense, fully connected layer with softmax activation. The proposed Conv1D+BiGRU+Attention model is efficient at classifying emotions in the Bengali music dataset than baseline methods. For our Bengali music dataset, the performance of our proposed model is 95%.

• Journal of Information Processing Systems
• /
• v.19 no.5
• /
• pp.576-589
• /
• 2023

To upgrade home style recommendations and user satisfaction, this paper proposes a personalized and optimized recommendation algorithm for interior design style based on local social network, which includes data acquisition by three-dimensional (3D) model, home-style feature definition, and style association mining. Through the analysis of user behaviors, the user interest model is established accordingly. Combined with the location-based social network of association rule mining algorithm, the association analysis of the 3D model dataset of interior design style is carried out, so as to get relevant home-style recommendations. The experimental results show that the proposed algorithm can complete effective analysis of 3D interior home style with the recommendation accuracy of 82% and the recommendation time of 1.1 minutes, which indicates excellent application effect.

• International Journal of Highway Engineering
• /
• v.18 no.2
• /
• pp.91-101
• /
• 2016

PURPOSES : The aim of this study was to compare the performance of the QUEENSOD method and the gravity model in estimating Origin-Destination (O/D) tables for a large-sized microscopic traffic simulation network. METHODS : In this study, an expressway network was simulated using the microscopic traffic simulation model, VISSIM. The gravity model and QUEENSOD method were used to estimate the O/D pairs between internal and between external zones. RESULTS: After obtaining estimations of the O/D table by using both the gravity model and the QUEENSOD method, the value of the root mean square error (RMSE) for O/D pairs between internal zones were compared. For the gravity model and the QUEENSOD method, the RMSE obtained were 386.0 and 241.2, respectively. The O/D tables estimated using both methods were then entered into the VISSIM networks and calibrated with measured travel time. The resulting estimated travel times were then compared. For the gravity model and the QUEENSOD method, the estimated travel times showed 1.16% and 0.45% deviation from the surveyed travel time, respectively. CONCLUSIONS : In building a large-sized microscopic traffic simulation network, an O/D matrix is essential in order to produce reliable analysis results. When link counts from diverse ITS facilities are available, the QUEENSOD method outperforms the gravity model.

• Journal of Korea Multimedia Society
• /
• v.23 no.3
• /
• pp.402-411
• /
• 2020

Power line noise in electrocardiogram signals makes it difficult to diagnose cardiovascular disease. ECG signals without power line noise are needed to increase the accuracy of diagnosis. In this paper, it is proposed DNN(Deep Neural Network) model to remove the power line noise in ECG. The proposed model is learned with noisy ECG, and clean ECG. Performance of the proposed model were performed in various environments(varying amplitude, frequency change, real-time amplitude change). The evaluation used signal-to-noise ratio and root mean square error (RMSE). The difference in evaluation metrics between the noisy ECG signals and the de-noising ECG signals can demonstrate effectiveness as the de-noising model. The proposed DNN model learning result was a decrease in RMSE 0.0224dB and a increase in signal-to-noise ratio 1.048dB. The results performed in various environments showed a decrease in RMSE 1.7672dB and a increase in signal-to-noise ratio 15.1879dB in amplitude changes, a decrease in RMSE 0.0823dB and a increase in signal-to-noise ratio 4.9287dB in frequency changes. Finally, in real-time amplitude changes, RMSE was decreased 0.3886dB and signal-to-noise ratio was increased 11.4536dB. Thus, it was shown that the proposed DNN model can de-noise power line noise in ECG.

• Korean Journal of Radiology
• /
• v.22 no.3
• /
• pp.476-488
• /
• 2021

Objective: We aimed to develop a deep neural network for segmenting lung parenchyma with extensive pathological conditions on non-contrast chest computed tomography (CT) images. Materials and Methods: Thin-section non-contrast chest CT images from 203 patients (115 males, 88 females; age range, 31-89 years) between January 2017 and May 2017 were included in the study, of which 150 cases had extensive lung parenchymal disease involving more than 40% of the parenchymal area. Parenchymal diseases included interstitial lung disease (ILD), emphysema, nontuberculous mycobacterial lung disease, tuberculous destroyed lung, pneumonia, lung cancer, and other diseases. Five experienced radiologists manually drew the margin of the lungs, slice by slice, on CT images. The dataset used to develop the network consisted of 157 cases for training, 20 cases for development, and 26 cases for internal validation. Two-dimensional (2D) U-Net and three-dimensional (3D) U-Net models were used for the task. The network was trained to segment the lung parenchyma as a whole and segment the right and left lung separately. The University Hospitals of Geneva ILD dataset, which contained high-resolution CT images of ILD, was used for external validation. Results: The Dice similarity coefficients for internal validation were 99.6 ± 0.3% (2D U-Net whole lung model), 99.5 ± 0.3% (2D U-Net separate lung model), 99.4 ± 0.5% (3D U-Net whole lung model), and 99.4 ± 0.5% (3D U-Net separate lung model). The Dice similarity coefficients for the external validation dataset were 98.4 ± 1.0% (2D U-Net whole lung model) and 98.4 ± 1.0% (2D U-Net separate lung model). In 31 cases, where the extent of ILD was larger than 75% of the lung parenchymal area, the Dice similarity coefficients were 97.9 ± 1.3% (2D U-Net whole lung model) and 98.0 ± 1.2% (2D U-Net separate lung model). Conclusion: The deep neural network achieved excellent performance in automatically delineating the boundaries of lung parenchyma with extensive pathological conditions on non-contrast chest CT images.

• Journal of Korea Water Resources Association
• /
• v.51 no.4
• /
• pp.347-354
• /
• 2018

In urban areas, runoff flow is drained through sewer networks as well as surface areas. Therefore, it is very important to consider sewer networks as a component of hydrological drainage processes when conducting urban inundation modelling. However, most researchers who have implemented urban inundation/flood modelling, instinctively simplified the sewer networks without the appropriate criteria. In this research, a 1D-2D fully coupled urban inundation model is applied to estimate the influence of sewer network simplification on urban inundation modelling based on the dendritic network classification. The one-dimensional (1D) sewerage system analysis model, which was introduced by Lee et al. (2017), is used to simulate inlet and overflow phenomena by interacting with surface flow. Two-dimensional (2D) unstructured meshes are also applied to simulate surface flow and are combined with the 1D sewerage analysis model. Sewer network pipes are simplified based on the dendritic network classification method, namely the second and third order, and all cases of pipes are conducted as a control group. Each classified network case, including a control group, is evaluated through their application to the 27 July 2011 extreme rainfall event, which caused severe inundation damages in the Sadang area in Seoul, South Korea. All cases are compared together regarding inundation area, inflow discharge and overflow discharge. Finally, relevant criterion for the simplification method is recommended.

• Journal of the Korean Society of Propulsion Engineers
• /
• v.25 no.5
• /
• pp.18-26
• /
• 2021

In this work, we have developed a 1D network model aimed at predicting eigenvalues for resonance frequency analysis in a lab-scale industrial gas turbine single nozzle combustion system. Modern industrial gas turbines generally adopt combustors with very complex geometry and flow path to meet various design requirements simultaneously. The current study has developed a network model for combustion systems with backflow at the same axial location. The modeling results of resonance frequencies and mode distributions for a given system using the network model were validated from comparisons with prediction results using a 3D Helmholtz solver.

• Journal of the Korea Convergence Society
• /
• v.11 no.9
• /
• pp.21-29
• /
• 2020

The Topic Modeling research, the methodology for deduction keyword within literature, has become active with the explosion of data from digital society transition. The research objective is to investigate research trends in D.N.A.(Data, Network, Artificial Intelligence) field using DTM(Dynamic Topic Model). DTM model was applied to the 1,519 of research projects with SW·A.I technology classifications among ICT(Information and Communication Technology) field projects between 6 years(2015~2020). As a result, technology keyword for D.N.A. field; Big data, Cloud, Artificial Intelligence, extended keyword; Unstructured, Edge Computing, Learning, Recognition was appeared every year, and accordingly that the above technology is being researched inclusively from other projects can be inferred. Finally, it is expected that the result from this paper become useful for future policy·R&D planning and corporation's technology·marketing strategy.