• Geophysics and Geophysical Exploration
• /
• v.24 no.2
• /
• pp.53-66
• /
• 2021

Velocity model building is an essential procedure in seismic data processing. Conventional techniques, such as traveltime tomography or velocity analysis take longer computational time to predict a single velocity model and the quality of the inversion results is highly dependent on human expertise. Full-waveform inversions also depend on an accurate initial model. Recently, deep neural network techniques are gaining widespread acceptance due to an increase in their integration to solving complex and nonlinear problems. This study investigated cases of seismic velocity model building using deep neural network techniques by classifying items according to the neural networks used in each study. We also included cases of generating training synthetic velocity models. Deep neural networks automatically optimize model parameters by training neural networks from large amounts of data. Thus, less human interaction is involved in the quality of the inversion results compared to that of conventional techniques and the computational cost of predicting a single velocity model after training is negligible. Additionally, unlike full-waveform inversions, the initial velocity model is not required. Several studies have demonstrated that deep neural network techniques achieve outstanding performance not only in computational cost but also in inversion results. Based on the research results, we analyzed and discussed the characteristics of deep neural network techniques for building velocity models.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.23 no.4
• /
• pp.415-423
• /
• 2021

Recently, an R-based point time series data validation system has been established for the statistical post processing and improvement of the National Center for AgroMeteorology-Land Atmosphere Modeling Package (NCAM-LAMP) medium-range prediction data. The time series verification system was used to compare the NCAM-LAMP with the AWS observations and GDAPS medium-range prediction model data operated by Korea Meteorological Administration. For this comparison, the model latitude and longitude data closest to the observation station were extracted and a total of nine points were selected. For each point, the characteristics of the model prediction error were obtained by comparing the daily average of the previous prediction data of air temperature, wind speed, and hourly precipitation, and then we tried to improve the next prediction data using Support Vector Machine( SVM) method. For three months from August to October 2017, the SVM method was used to calibrate the predicted time series data for each run. It was found that The SVM-based correction was promising and encouraging for wind speed and precipitation variables than for temperature variable. The correction effect was small in August but considerably increased in September and October. These results indicate that the SVM method can contribute to mitigate the gradual degradation of medium-range predictability as the model boundary data flows into the model interior.

• Journal of the Korea Society for Simulation
• /
• v.31 no.1
• /
• pp.11-18
• /
• 2022

Recently, research to classify human activity using imagery has been actively conducted for the purpose of crime prevention and facility safety in public places and facilities. In order to improve the performance of human activity classification, most studies have applied deep learning based-transfer learning. However, despite the increase in the number of backbone network models that are the basis of deep learning as well as the diversification of architectures, research on finding a backbone network model suitable for the purpose of operation is insufficient due to the atmosphere of using a certain model. Thus, this study applies the transfer learning into recently developed deep learning backborn network models to build an intelligent system that classifies human activity using imagery. For this, 12 types of active and high-contact human activities based on sports, not basic human behaviors, were determined and 7,200 images were collected. After 20 epochs of transfer learning were equally applied to five backbone network models, we quantitatively analyzed them to find the best backbone network model for human activity classification in terms of learning process and resultant performance. As a result, XceptionNet model demonstrated 0.99 and 0.91 in training and validation accuracy, 0.96 and 0.91 in Top 2 accuracy and average precision, 1,566 sec in train process time and 260.4MB in model memory size. It was confirmed that the performance of XceptionNet was higher than that of other models.

• Smart Media Journal
• /
• v.10 no.4
• /
• pp.54-62
• /
• 2021

As the volume of online e-commerce transactions increases, items diversify and the classification becomes complicated. E-commerce platforms that specialize in dealing only in one area are emerging, and the area is diversifying. Three problems were identified by researching the craft online e-commerce platform, one of the various types of professional e-commerce platforms. First of all, although craft materials are diversified and complex on the platform, the existing craft e-commerce system is fragmented in structure to categorize complex crafts, making it difficult to accurately present search results that meet various criteria. Second, although appearance is the main reason for purchasing artifacts, it is rare for users to categorize them according to appearance, so they have to judge and filter each work directly. Finally, the language entered when searching for artifacts by non-technical experts is not reflected in the language used to categorize artifacts in the taxonomic system, so the language used for searching is highly accurate. Therefore, the purpose of this study is to add and consider complex attributes in the field of technology to meet the search criteria. Properties to be added must include the main appearance in the search for artifacts. In addition, the government aims to develop a taxonomic system that can reflect non-experts' search languages in the search of works through artificial intelligence natural language processing technology.

• Journal of the Korean Geotechnical Society
• /
• v.37 no.11
• /
• pp.119-129
• /
• 2021

The process of constructing the DB in the current geotechnical information DB system needs a lot of human and time resource consumption. In addition, it causes accuracy problems frequently because the current input method is a person viewing the PDF and directly inputting the results. Therefore, this study proposes building an automatic digital DB using AI (artificial intelligence) of boring logs. In order to automatically construct DB for various boring log formats without exception, the boring log forms were classified using the deep learning model ResNet 34 for a total of 6 boring log forms. As a result, the overall accuracy was 99.7, and the ROC_AUC score was 1.0, which separated the boring log forms with very high performance. After that, the text in the PDF is automatically read using the robotic processing automation technique fine-tuned for each form. Furthermore, the general information, strata information, and standard penetration test information were extracted, separated, and saved in the same format provided by the geotechnical information DB system. Finally, the information in the boring log was automatically converted into a DB at a speed of 140 pages per second.

• The Journal of the Korea Contents Association
• /
• v.22 no.2
• /
• pp.10-21
• /
• 2022

In order to improve formal presentation attitudes such as presentation of job interviews and presentation of project results at the company, there are few automated methods other than observation by colleagues or professors. In previous studies, it was reported that the speaker's stable speech and gaze processing affect the delivery power in the presentation. Also, there are studies that show that proper feedback on one's presentation has the effect of increasing the presenter's ability to present. In this paper, considering the positive aspects of correction, we developed a program that intelligently corrects the wrong presentation habits and attitudes of college students through facial analysis of videos and analyzed the proposed program's performance. The proposed program was developed through web-based verification of the use of redundant words and facial recognition and textualization of the presentation contents. To this end, an artificial intelligence model for classification was developed, and after extracting the video object, facial feature points were recognized based on the coordinates. Then, using 4000 facial data, the performance of the algorithm in this paper was compared and analyzed with the case of facial recognition using a Teachable Machine. Use the program to help presenters by correcting their presentation attitude.

• Journal of the Korea Convergence Society
• /
• v.12 no.5
• /
• pp.23-29
• /
• 2021

Recent deep learning-based natural language processing studies are conducting research to improve performance by training large amounts of data from various sources together. However, there is a possibility that the methodology of learning by combining data from various sources into one may prevent performance improvement. In the case of machine translation, data deviation occurs due to differences in translation(liberal, literal), style(colloquial, written, formal, etc.), domains, etc. Combining these corpora into one for learning can adversely affect performance. In this paper, we propose a new Corpus Weight Balance(CWB) method that considers the balance between parallel corpora in machine translation. As a result of the experiment, the model trained with balanced corpus showed better performance than the existing model. In addition, we propose an additional corpus construction process that enables coexistence with the human translation market, which can build high-quality parallel corpus even with a monolingual corpus.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.4
• /
• pp.11-19
• /
• 2021

An abnormal object refers to a person, an object, or a mechanical device that performs abnormal and unusual behavior and needs observation or supervision. In order to detect this through artificial intelligence algorithm without continuous human intervention, a method of observing the specificity of temporal features using optical flow technique is widely used. In this study, an abnormal situation is identified by learning an algorithm that translates an input image frame to an optical flow image using a Generative Adversarial Network (GAN). In particular, we propose a technique that improves the pre-processing process to exclude unnecessary outliers and the post-processing process to increase the accuracy of identification in the test dataset after learning to improve the performance of the model's abnormal behavior identification. UCSD Pedestrian and UMN Unusual Crowd Activity were used as training datasets to detect abnormal behavior. For the proposed method, the frame-level AUC 0.9450 and EER 0.1317 were shown in the UCSD Ped2 dataset, which shows performance improvement compared to the models in the previous studies.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.26 no.3
• /
• pp.11-20
• /
• 2022

The main purpose of this study is to investigate the applicability of the shear wave tomography technology as a non-destructive testing method to evaluate the debonding between the track concrete layer (TCL) and the hydraulically stabilized based course (HSB) of concrete slab tracks for the Korea high-speed railway system. A commercially available multi-channel shear wave measurement device (MIRA) is used to evaluate debonding defects in full-scaled mock-up test specimen that was designed and constructed according to the Rheda 200 system. A part of the mock-up specimen includes two artificial debonding defects with a length and a width of 400mm and thicknesses of 5mm and 10mm, respectively. The tomography images obtained by a MIRA on the surface of the concrete specimens are effective for visualizing the debonding defects in concrete. In this study, a simple image processing method is proposed to suppress the noisy signals reflected from the embedded items (reinforcing steel, precast sleeper, insert, etc.) in TCL, which significantly improves the readability of debonding defects in shear wave tomography images. Results show that debonding maps constructed in this study are effective for visualizing the spatial distribution and the depths of the debondiing defects in the railway concrete slab specimen.

• Geophysics and Geophysical Exploration
• /
• v.25 no.1
• /
• pp.14-25
• /
• 2022

The passive surface wave method measures seismic signals from ambient noises or vibrations of natural phenomena without using an artificial source. Since passive sources are usually in lower frequencies than artificial ones being able to ensure the information on deeper geological structures, the passive surface wave method can investigate deeper geological structures. In the passive method, frequency dispersion curves are obtained after data acquisition, and the dispersion curves are analyzed by assuming 1D-layered earth, which is like the method of active surface wave survey. However, when computing dispersion curves, the passive method first obtains and analyzes coherence curves of received signals from a set of receivers based on spatial autocorrelation. In this review, we explain how passive surface wave methods measure signals, and make data processing and interpretation, before analyzing field application cases.