• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.7
• /
• pp.303-314
• /
• 2023

Voice conversion, a technology that allows an individual's speech data to be regenerated with the acoustic properties(tone, cadence, gender) of another, has countless applications in education, communication, and entertainment. This paper proposes an approach based on the StarGAN-VC model that generates realistic-sounding speech without requiring parallel utterances. To overcome the constraints of the existing StarGAN-VC model that utilizes one-hot vectors of original and target speaker information, this paper extracts feature vectors of target speakers using a pre-trained version of Rawnet3. This results in a latent space where voice conversion can be performed without direct speaker-to-speaker mappings, enabling an any-to-any structure. In addition to the loss terms used in the original StarGAN-VC model, Wasserstein distance is used as a loss term to ensure that generated voice segments match the acoustic properties of the target voice. Two Time-Scale Update Rule (TTUR) is also used to facilitate stable training. Experimental results show that the proposed method outperforms previous methods, including the StarGAN-VC network on which it was based.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.10
• /
• pp.27-35
• /
• 2023

In this paper, we propose a process of increasing productivity by applying a deep learning-based defect detection and classification system to the prepreg fiber manufacturing process, which is in high demand in the field of producing composite materials. In order to apply it to toe prepreg manufacturing equipment that requires a solution due to the occurrence of a large amount of defects in various conditions, the optimal environment was first established by selecting cameras and lights necessary for defect detection and classification model production. In addition, data necessary for the production of multiple classification models were collected and labeled according to normal and defective conditions. The multi-classification model is made based on CNN and applies pre-learning models such as VGGNet, MobileNet, ResNet, etc. to compare performance and identify improvement directions with accuracy and loss graphs. Data augmentation and dropout techniques were applied to identify and improve overfitting problems as major problems. In order to evaluate the performance of the model, a performance evaluation was conducted using the confusion matrix as a performance indicator, and the performance of more than 99% was confirmed. In addition, it checks the classification results for images acquired in real time by applying them to the actual process to check whether the discrimination values are accurately derived.

• Journal of IKEEE
• /
• v.28 no.2
• /
• pp.150-157
• /
• 2024

In this paper, we propose a novel algorithm for predicting the number of apples on an apple tree using a deep learning-based object detection model and a polynomial regression model. Measuring the number of apples on an apple tree can be used to predict apple yield and to assess losses for determining agricultural disaster insurance payouts. To measure apple fruit load, we photographed the front and back sides of apple trees. We manually labeled the apples in the captured images to construct a dataset, which was then used to train a one-stage object detection CNN model. However, when apples on an apple tree are obscured by leaves, branches, or other parts of the tree, they may not be captured in images. Consequently, it becomes difficult for image recognition-based deep learning models to detect or infer the presence of these apples. To address this issue, we propose a two-stage inference process. In the first stage, we utilize an image-based deep learning model to count the number of apples in photos taken from both sides of the apple tree. In the second stage, we conduct a polynomial regression analysis, using the total apple count from the deep learning model as the independent variable, and the actual number of apples manually counted during an on-site visit to the orchard as the dependent variable. The performance evaluation of the two-stage inference system proposed in this paper showed an average accuracy of 90.98% in counting the number of apples on each apple tree. Therefore, the proposed method can significantly reduce the time and cost associated with manually counting apples. Furthermore, this approach has the potential to be widely adopted as a new foundational technology for fruit load estimation in related fields using deep learning.

• Journal of Advanced Navigation Technology
• /
• v.27 no.6
• /
• pp.710-715
• /
• 2023

Urbanization brings many challenges such as traffic, housing, and environment. To solve these problems, researchers are working on new transportation systems like urban air mobility (UAM). UAM aircraft should fly safely without burdening the existing air traffic system in the early stage of low-density operation. The airspace should also be managed and operated efficiently. Therefore it is important to make urban air traffic predictable by using corridors and collecting data on low-density operations in the early stage. For this purpose various simulations are needed before operation to create scenarios that estimate potential collisions between UAM aircraft and to evaluate the risks of aircraft spacing, loss of separation (LoS), and near mid air collision (NMAC). This paper focuses on identifying the requirements and considerations for setting separation standards for urban air traffic based on the results of studies.

• The Korean Journal of Mycology
• /
• v.29 no.1
• /
• pp.15-21
• /
• 2001

This study was conducted to investigate the immunomodulatory activities of proteoglycan extracted from cultured mycelia of Ganoderma lucidum IY009. The proteoglycan contained two polymer peaks, one was the higher MW peak of 2,000 kD and the other was low peaks of 12kD. To understand the part of strong pharmaceutical activity between two peak, the proteoglycan was separated by ultrafiltration and column chromatography and then examined the various pharmaceutical effects. High molecular weight fraction possesing high content of ${\beta}-linked$ glucan was exhibited high antitumor activity, against sarcoma 180 bearing ICR mouse. And also, anticomplementary activity was highly observed in high molecule fraction than low it fraction. When the raw 264.7 and murine peritoneal macrophage treated with low fraction, high fraction and other stimuli. The activities inducing tumor necrosis factor of the high factions were $2.2{\sim}2.5$ times stronger than that of low fraction.

• The Korean Journal of Nuclear Medicine
• /
• v.30 no.4
• /
• pp.548-559
• /
• 1996

A series of performance measurements of positron emission tomography (PET) were performed following the recommendations of the Computer and Instrumentation Council of the Society of Nuclear Medicine and the National Electrical Manufacturers Association. We investigated the performance of the General Electric $Advance^{TM}$ PET. The measurements include the basic intrinsic tests of spatial resolution, scatter fraction, sensitivity, and count rate losses and randoms. They also include the tests of the accuracy of corrections: count rate linearity correction, uniformity correction, scatter correction and attenuation correction. GE $Advance^{TM}$ PET has bismuth germanate oxide crystals (4.0mm transaxial ${\times}$ 8.1mm axial ${\times}$ 30.0mm radial) in 18 rings, which form 35 imaging planes spaced by 4.25mm. The system has retractable tungsten septa 1mm thick and 12cm long. Transaxial resolution was 4.92mm FWHM in 2D and 5.14mm FWHM in 3D at the center. Average axial resolution in 2D decreased from 3.91mm FWHM at the center to 6.49mm FWHM at R=20cm. Average scatter fraction of direct and cross slices was 9.57%. Dead-time losses of 50% corresponded to a radioactivity concentration of $4.86{\mu}Ci/cc$ and a true count rate of 519 kcps in 2D. The accuracy of count rate linearity correction was 1.84% at the activity of $4.50{\mu}Ci/cc$. Non-uniformity was 2.06% in 2D and 2.93% in 3D. Remnant errors after scatter correction were 0.55% in 2D and 4.12% in 3D. The errors of attenuation correction were 6.21% (air), 0.20% (water), -6.32% (teflon) in 2D and 5.00% (air), 6.94% (water), 3.01% (teflon) in 3D. The results indicate the performance of GE $Advance^{TM}$ PET scanner to be well suited for clinical and research applications.

• Journal of Korea Spatial Information System Society
• /
• v.5 no.1 s.9
• /
• pp.49-63
• /
• 2003

Recently, two international standard organizations, ISO and OGC, have done the work of standardization for GIS. Current standardization work for providing interoperability among GIS DB focuses on the design of open interfaces. But, this work has not considered procedures and methods for designing GIS DB. Eventually, GIS DB has its own model. When we share the data by open interface among heterogeneous GIS DB, differences between models result in the loss of information. Our aim in this paper is to revise the design guidelines for geographic information databases in order to make consistent spatial data models, logical structures, and semantic structure of populated geographical databases. In details, we propose standard guidelines which convert ISO abstract schema into relation model, object-relation model, object-centered model, and geometry-centered model. Furthermore, we provide sample models for applying these guidelines in commercial GIS S/Ws. Building GIS DB based on design guidelines proposed in the paper has the following advantages: the interoperability among databases, the standardization of schema definitions, and the catalogue of GIS databases through.

• Journal of Internet Computing and Services
• /
• v.16 no.2
• /
• pp.77-83
• /
• 2015

Frequent pattern mining, which is one of the major areas actively studied in data mining, is a method for extracting useful pattern information hidden from large data sets or databases. Moreover, frequent pattern mining approaches have been actively employed in a variety of application fields because the results obtained from them can allow us to analyze various, important characteristics within databases more easily and automatically. However, traditional frequent pattern mining methods, which simply extract all of the possible frequent patterns such that each of their support values is not smaller than a user-given minimum support threshold, have the following problems. First, traditional approaches have to generate a numerous number of patterns according to the features of a given database and the degree of threshold settings, and the number can also increase in geometrical progression. In addition, such works also cause waste of runtime and memory resources. Furthermore, the pattern results excessively generated from the methods also lead to troubles of pattern analysis for the mining results. In order to solve such issues of previous traditional frequent pattern mining approaches, the concept of representative pattern mining and its various related works have been proposed. In contrast to the traditional ones that find all the possible frequent patterns from databases, representative pattern mining approaches selectively extract a smaller number of patterns that represent general frequent patterns. In this paper, we describe details and characteristics of pattern condensing techniques that consider the maximality or closure property of generated frequent patterns, and conduct comparison and analysis for the techniques. Given a frequent pattern, satisfying the maximality for the pattern signifies that all of the possible super sets of the pattern must have smaller support values than a user-specific minimum support threshold; meanwhile, satisfying the closure property for the pattern means that there is no superset of which the support is equal to that of the pattern with respect to all the possible super sets. By mining maximal frequent patterns or closed frequent ones, we can achieve effective pattern compression and also perform mining operations with much smaller time and space resources. In addition, compressed patterns can be converted into the original frequent pattern forms again if necessary; especially, the closed frequent pattern notation has the ability to convert representative patterns into the original ones again without any information loss. That is, we can obtain a complete set of original frequent patterns from closed frequent ones. Although the maximal frequent pattern notation does not guarantee a complete recovery rate in the process of pattern conversion, it has an advantage that can extract a smaller number of representative patterns more quickly compared to the closed frequent pattern notation. In this paper, we show the performance results and characteristics of the aforementioned techniques in terms of pattern generation, runtime, and memory usage by conducting performance evaluation with respect to various real data sets collected from the real world. For more exact comparison, we also employ the algorithms implementing these techniques on the same platform and Implementation level.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.24 no.6
• /
• pp.617-628
• /
• 2022

A ground support pattern should be designed by properly integrating various support materials in accordance with the rock mass grade when constructing a tunnel, and a technical decision must be made in this process by professionals with vast construction experiences. However, designing supports at the early stage of tunnel design, such as feasibility study or basic design, may be very challenging due to the short timeline, insufficient budget, and deficiency of field data. Meanwhile, the design of the support pattern can be performed more quickly and reliably by utilizing the machine learning technique and the accumulated design data with the rapid increase in tunnel construction in South Korea. Therefore, in this study, the design data and ground exploration data of 48 road tunnels in South Korea were inspected, and data about 19 items, including eight input items (rock type, resistivity, depth, tunnel length, safety index by tunnel length, safety index by rick index, tunnel type, tunnel area) and 11 output items (rock mass grade, two items for shotcrete, three items for rock bolt, three items for steel support, two items for concrete lining), were collected to automatically determine the rock mass class and the support pattern. Three machine learning models (S1, A1, A2) were developed using two machine learning algorithms (SVR, ANN) and organized data. As a result, the A2 model, which applied different loss functions according to the output data format, showed the best performance. This study confirms the potential of support pattern design using machine learning, and it is expected that it will be able to improve the design model by continuously using the model in the actual design, compensating for its shortcomings, and improving its usability.

• Journal of Bio-Environment Control
• /
• v.32 no.4
• /
• pp.434-441
• /
• 2023

Determining the size or area of a plant's leaves is an important factor in predicting plant growth and improving the productivity of indoor farms. In this study, we developed a convolutional neural network (CNN)-based model to accurately predict the length and width of lettuce leaves using photographs of the leaves. A callback function was applied to overcome data limitations and overfitting problems, and K-fold cross-validation was used to improve the generalization ability of the model. In addition, ImageDataGenerator function was used to increase the diversity of training data through data augmentation. To compare model performance, we evaluated pre-trained models such as VGG16, Resnet152, and NASNetMobile. As a result, NASNetMobile showed the highest performance, especially in width prediction, with an R_squared value of 0.9436, and RMSE of 0.5659. In length prediction, the R_squared value was 0.9537, and RMSE of 0.8713. The optimized model adopted the NASNetMobile architecture, the RMSprop optimization tool, the MSE loss functions, and the ELU activation functions. The training time of the model averaged 73 minutes per Epoch, and it took the model an average of 0.29 seconds to process a single lettuce leaf photo. In this study, we developed a CNN-based model to predict the leaf length and leaf width of plants in indoor farms, which is expected to enable rapid and accurate assessment of plant growth status by simply taking images. It is also expected to contribute to increasing the productivity and resource efficiency of farms by taking appropriate agricultural measures such as adjusting nutrient solution in real time.