• KIPS Transactions on Software and Data Engineering
• /
• v.8 no.3
• /
• pp.115-122
• /
• 2019

For Korean phoneme recognition, Hidden Markov-Gaussian Mixture model(HMM-GMM) or hybrid models which combine artificial neural network with HMM have been mainly used. However, current approach has limitations in that such models require force-aligned corpus training data that is manually annotated by experts. Recently, researchers used neural network based phoneme recognition model which combines recurrent neural network(RNN)-based structure with connectionist temporal classification(CTC) algorithm to overcome the problem of obtaining manually annotated training data. Yet, in terms of implementation, these RNN-based models have another difficulty in that the amount of data gets larger as the structure gets more sophisticated. This problem of large data size is particularly problematic in the Korean language, which lacks refined corpora. In this study, we introduce CTC algorithm that does not require force-alignment to create a Korean phoneme recognition model. Specifically, the phoneme recognition model is based on convolutional neural network(CNN) which requires relatively small amount of data and can be trained faster when compared to RNN based models. We present the results from two different experiments and a resulting best performing phoneme recognition model which distinguishes 49 Korean phonemes. The best performing phoneme recognition model combines CNN with 3hop Bidirectional LSTM with the final Phoneme Error Rate(PER) at 3.26. The PER is a considerable improvement compared to existing Korean phoneme recognition models that report PER ranging from 10 to 12.

• The Journal of the Korea Contents Association
• /
• v.21 no.4
• /
• pp.785-794
• /
• 2021

This study investigated the current status of the design and direction of improvement of the sign system of urban railway stations in the Seoul metropolitan area to see how difficult it is for users to use. As a research method, first, a theoretical study on the literature of the urban railway sign system, second, a survey on the current status of the sign design of the Gimpo Goldline in the Seoul metropolitan area, and third, a survey on the actual use of the sign and problems was conducted. As a result of the current status survey, the location sign is in central alignment with the font of the guide sign, and the guidance sign is in front/back alignment according to the location of the entrance and exit. The background of the font is white and the text is dark black to make it look good to the user. On the other hand, guidance sign is written in white color and background color is gray color. Currently, the Gimpo Urban Railway Guidance Sign System is newly produced and installed, but there is a lack of user design and needs design maintenance that can clearly deliver information to users.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.22 no.2
• /
• pp.378-385
• /
• 2021

The production of plum jangachi requires the cleaning of plums, immersion in saltwater, separation of seed and pulp, removal of impurities, and addition of sugar. In most farms, the separation of plum seed and pulp is carried out manually, requiring considerable labor, which is why plum jangachi is expensive. To solve this problem, this study designed and manufactured automatic, semi-automatic plum seed and pulp separators. During the design process, the characteristics were compared, and the machine power was determined through on-site test after manufacture. As a result, automatic machines used plums 180° arrayed and six reverse-edged blades, semi-automated plums 180° arrayed, and six blades, each with a 68% and 57% pulp recovery rate and a machine power of 80 kg/h and 62 kg/h respectively. Overall, the mechanization of plum processed food will reduce labor and increase the market value of plums compared to the previous method.

• The KIPS Transactions:PartB
• /
• v.15B no.4
• /
• pp.275-284
• /
• 2008

With increases in recent security requirements, biometric technology such as fingerprints, faces and iris recognitions have been widely used in many applications including door access control, personal authentication for computers, internet banking, automatic teller machines and border-crossing controls. Finger vein recognition uses the unique patterns of finger veins in order to identify individuals at a high level of accuracy. This paper proposes new device and methods for touchless finger vein recognition. This research presents the following five advantages compared to previous works. First, by using a minimal guiding structure for the finger tip, side and the back of finger, we were able to obtain touchless finger vein images without causing much inconvenience to user. Second, by using a hot mirror, which was slanted at the angle of 45 degrees in front of the camera, we were able to reduce the depth of the capturing device. Consequently, it would be possible to use the device in many applications having size limitations such as mobile phones. Third, we used the holistic texture information of the finger veins based on a LBP (Local Binary Pattern) without needing to extract accurate finger vein regions. By using this method, we were able to reduce the effect of non-uniform illumination including shaded and highly saturated areas. Fourth, we enhanced recognition performance by excluding non-finger vein regions. Fifth, when matching the extracted finger vein code with the enrolled one, by using the bit-shift in both the horizontal and vertical directions, we could reduce the authentic variations caused by the translation and rotation of finger. Experimental results showed that the EER (Equal Error Rate) was 0.07423% and the total processing time was 91.4ms.

• Journal of the Korean Society of Radiology
• /
• v.17 no.3
• /
• pp.305-314
• /
• 2023

With the development of medical technology and radiation treatment equipment, the frequency of high-precision radiation therapy such as intensity modulation radiation therapy has increased. Image-guided radiation therapy has become essential for radiation therapy in precise and complex treatment plans. In particular, with the introduction of imaging equipment for diagnosis in a linear accelerator, CBCT scanning became possible, which made it possible to calibrate and correct the patient's posture through 3D images. Although more precise reproduction of the patient's posture has become possible, the exposure dose delivered to the patient during the image acquisition process cannot be ignored. Radiation optimization is necessary in the field of radiation therapy, and efforts to reduce exposure are necessary. However, when acquiring 3D CBCT images by changing the imaging conditions to reduce exposure, there should be no image quality or artefacts that would make it impossible to align the patient's position. In this study, Rando phantom was used to scan and evaluate images for each shooting condition. The highest SNR was obtained at 100 kV 80 mA 25 ms F1 filter 180°. As the tube voltage and tube current increased, the noise decreased, and the bowtie filter showed the optimal effect at high tube current. Based on the actual scanned images, it was confirmed that patient alignment was possible under all imaging conditions, and that image-guided radiation therapy for patient alignment was possible under the condition of 70 kV 10 mA 20 ms F0 filter 180°, which showed the lowest SNR. In this study, image evaluation was conducted according to the imaging conditions, and low tube voltage, tube current, and small rotation angle scan are expected to be effective in reducing radiation exposure. Based on this, the patient's exposure dose should be kept as low as possible during CBCT imaging.

• Journal of Biomedical Engineering Research
• /
• v.25 no.6
• /
• pp.531-536
• /
• 2004

In fracture treatment with external fixators, the inter-fragmentary movements at the fracture site affect the fracture healing process, and these movements are highly related to the stiffness of external fixation systems. Therefore, in order to provide the optimal fracture healing at the fracture site, it is essential to understand the relationship between the stiffness and the system configurations in external fixation system. In this study we investigated the influences of system configuration parameters on the stiffness in the finite element analysis of an external fixation system of a long bone. The system alignment, the geometric and the material non-linearity of the pin, the joint stiffness and the callus formation were considered in the finite element model. In the first, the system stiffness of the developed finite element model was compared with the experiment data for model validation. The consideration of the joint stiffness and nonlinearity of the model improved the system stiffness results. The joint stiffness, the non-alignment of the system decreased the system stiffness while the callus formation increased the system stiffness. The present results provided the biomechanical basis of rational guidelines for design improvements of external fixators and pre-op. planning to maximize the system stiffness in fracture surgery.

• Journal of Life Science
• /
• v.19 no.9
• /
• pp.1218-1225
• /
• 2009

Difractose anhydrides (DFAs) is studied as a sweetener for diabetics because of its structural property. DFAs have four types: DFA I, III, IV (degradation of levan) and V (degradation of inulin). Especially, DFA IV has been shown to enhance the absorption of calcium in experiments using rats. Levan fructotransferase is an enzyme for producing di-d-fructose-2,6':6,2-dianhydride (DFA IV). To identify structural characterization, we purified wild-type and mutants (D63A, D195N and N85S) of levan fructotransferase (LFTase) from Microbacterium sp. AL-210. These proteins were purified to apparent homogeneity by Ni-NTA affinity column, Q-sepharose ion exchange and gel filtration chromatography and detected by SDS-PAGE. They were also analyzed by circular dichroism (CD) measurements, JNET secondary structure prediction, activity measurements at various temperatures, and pH analysis. The optimum pH for the enzyme-catalyzed reaction was pH 7.5 and optimum temperature was observed at $55^{\circ}C$. Along with wild-type LFTase, mutants were analyzed by CD measurement, fluorescence analysis and differential scanning calorimetry (DSC). N85S showed less $\alpha$-helix and more $\beta$ strand than others. Also, N85S showed almost the same curve as wild-type in their steady-state fluorescence spectra, whereas mutant D63A and D195N showed higher intensity than wild-type. The amino acid sequence of wild-type LFTase was compared to the sequences of exo-inulinase from Aspergillus awamori, a plant fructan 1-exohydrolase from Cichorium intybus, and Thermotogo maritime (Tm) invertase and showed a high identity with Exo-inulinase from Aspergillus awamori.

• Journal of Oral Medicine and Pain
• /
• v.38 no.1
• /
• pp.35-51
• /
• 2013

The aims of this study was to observed an effect of antioxidative in cardiac muscle of high fat diet induced obesity rat by treadmill exercise with intensity and time. Thirty-two Sprauge-Dawley rats which were divided into four group. Normal, Control(high fat diet induced obesity rat), Experimental I(high intensity intermittent exercise in high fat diet induced obesity rat), Experimental II(moderate intensity endurance exercise in high fat diet induced obesity rat). The results of this study were as follows: 1. In change of body weight, the outcome of each group significantly difference compared with control. Also, 1 to 3 weeks significantly different compared with pre valu experimental I and II(p<0.001). 2. In change of lipid profile, the outcome of each group significantly difference compared with control(p<0.001). Difference between experimental I and II is not significantly. 3. In change of antioxidative enzymes(SOD, CAT, GPx) in myocardium, there are significant difference between control and experimental II, and also between control and experimental I(p<0.001). 4. In change of antioxidative protein MCR-1, the outcome of each group significantly difference compared with control(p<0.01). Experimental II was most significantly difference than the other group(p<0.001). The above results suggest that treadmill exercise effectively reduced in fat. It would be considered that moderate intensity endurance exercise has an effects on improved antioxidative enzyme in cardiac muscle of high fat diet induced obesity rat.

• Journal of Aerospace System Engineering
• /
• v.11 no.4
• /
• pp.48-53
• /
• 2017

Liquid rocket engines (LREs) are very complex systems that include combustion chambers, turbopumps, gas generators, ducts and tubes, valves and etc. Most components of the LREs require higher than or equal to level 6 IT (ISO Tolerance). The components along with pipe line and/or tubing must dispose not to interfere each other. In addition, effectiveness of maintenance and service after assembling should be considered when the allocation of the components are determined. Especially at the stage of the development, tolerance accumulations or unpredictable errors may result in misalignment and/or mismatches at interfaces of the parts. Namely, it is the engine assembling process that many inherent risks are realized and crises or incidents occur. Therefore, a rapid reaction system should be prepared. In this research, 7 tonf class liquid rocket assembling process was studied and actual building steps were introduced.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.25 no.4A
• /
• pp.464-470
• /
• 2000

In this paper, we introduce a modified interference cancellation scheme to overcome MAI(Multiple Access Interference) in DS-CDMA. Among ICs(Interference Cancellers), PIC(Parallel Interference Canceller) requires the more stages to have the better BER(Bit Error Rate), and SIC(Successive Interference Canceller) faces the problems of power reordering and large delays. To resolve the problems of them, we propose a modified interference cancellation scheme combining SIC and PIC. The proposed interference canceller has improved the performance through the total received signal applied sorting method is fed into the initial stage. Its structure extracts the good characteristics because of the input signal removed MAI in the initial stage. The characteristics of the proposed system are as follows (1) smaller delay, and the strongest user signal is also reduced to MAI(unlike SIC), (2) it has the good BER performance only using fewer stages (unlike PIC).