• Annual Conference of KIPS
• /
• 2017.11a
• /
• pp.799-802
• /
• 2017

We attempt to combines concepts of Morphological Operator(MO) and Convolutional Neural Networks(CNN) to improve image-to-image translation. To do this, we propose an operation that approximates morphological operations. Also we propose S-Convolution, an operation that extends the operation to use multiple filters like CNN. The experiment result shows that it can learn MO with big filter using multiple S-convolution layer of small filter. To validate effectiveness of the proposed layer in image-to-image translation we experiment with GAN with S-convolution applied. The result showed that GAN with S-convolution can achieve distinct result from that of GAN with CNN.

• ETRI Journal
• /
• v.45 no.3
• /
• pp.462-478
• /
• 2023

In quantum computing, the use of Pauli frames through software traces of classical computers improves computation efficiency. In previous studies, error correction and Pauli operation tracking have been performed simultaneously using integrated Pauli frames in the physical layer. In such a complex processing structure, the number of simultaneous operations processed in the physical layer exponentially increases as the distance of the surface code encoding logical qubit increases. This study proposes a Pauli frame management architecture partitioned into two layers for a lattice surgery-based surface code and describes its structure and operation rules. To evaluate the effectiveness of our method, we generated a random circuit according to the gate ratios constituting the commonly known quantum circuits and compared the generated circuit with the existing Pauli frame and our method. Simulations show a decrease of about 5% over traditional methods. In the case of experiments that only increase the code distance of the logical qubit, it can be seen that the effect of reducing the physical operation through the logical Pauli frame becomes more important.

• Journal of Korean Neurosurgical Society
• /
• v.53 no.2
• /
• pp.125-128
• /
• 2013

A 65-year-old man who had lateral cervical disc herniation underwent cervical posterior laminoforaminotomy at C5-6 and C6-7 level right side. During the operation, there was no serious surgical bleeding event. After operation, he complained persistent right shoulder pain and neck pain. Repeated magnetic resonance image (MRI) showed diffuse cervical epidural hematoma (EDH) extending from C5 to T1 level right side and spinal cord compression at C5-6-7 level. He underwent exploration. There was active bleeding at muscular layer. Muscular active bleeding was controlled and intramuscular hematoma was removed. The patient's symptom was reduced after second operation. Symptomatic postoperative spinal EDH requiring reoperation is rare. Meticulous bleeding control is important before wound closure. In addition, if patient presents persistent or aggravated pain after operation, rapid evaluation using MRI and second look operation is needed as soon as possible.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.41 no.2
• /
• pp.101-106
• /
• 2005

This study was carried out to observe the underwater geometry of the anchovy boat seine by two fishing operation systems. One was a traditional operation system that used a fish detecting boat, another was a new operation system that used the same fishing gear with floats and sinkers in comparison with a fish detecting boat. Vertical opening of the fishing gear that used a casting net, working depth of each part of the fishing gear that used a new operation system was almost canstant, but was gradually shallow in a traditional operation system. Just before hauling net, working depth of each part of the fishing gear that used a new operation system was maintained stable, but was vertically unstable in a traditional operation system because as rear part of the fishing gear was risen up the upper layer. Just after hauling net, working depth of the fishing gear that used a new operation system was also maintained stable, and then anchovy school could be lead to bag net.

• Proceedings of the Korean Society for Technology of Plasticity Conference
• /
• 2008.10a
• /
• pp.401-404
• /
• 2008

In the present investigation, a high precision thermal imprint system for micro ceramic products was developed and the net-shape manufacturing of multi-layer ceramic reflector for LED (Light Emitting Diode) was conducted with a precision metal die. Workpiece used in the present investigation were the multi-layer laminated ceramic sheets with pre-punched holes. The cavity with arbitrary angle was formed on the circular and rectangular holes of the ceramic sheets. During the imprinting process, the ambient temperature of the imprint system was kept over the transition temperature of the ceramic sheet and then rapidly cooled. The results in this paper show that the present method can be successfully applied to the fabrication of very small size hole array for ceramic reflector in a one step operation.

• Journal of Mechanical Science and Technology
• /
• v.17 no.6
• /
• pp.896-905
• /
• 2003

A linear motor has many advantages next to conventional feed mechanisms: high transitional speed and acceleration, high control performance, and good positioning accuracy at high speed. Through the omission of a power transfer element, the linear motor shows no wear and no backlash, has a long lifetime, and is easy to assemble. A disadvantage of the linear motor is low efficiency and resultant high-temperature rise in itself and neighboring structures during operation. This paper presents the thermal behavior of the linear motor as a feed mechanism in machine tools. To improve the thermal behavior, an insulation layer is used. By placing the insulation layer between the primary part and the machine table, both the temperature difference and the temperature fluctuation in the machine table due to a varying motor load are reduced.

• Proceedings of the KIEE Conference
• /
• 2008.10b
• /
• pp.513-514
• /
• 2008

A scalable satellite broadcasting system is designed to guarantee quality of multimedia service cooperatively utilizing both Ku Band and Ka band. In the system, video streams are separated into a base layer and an enhancement layer, and are transmitted over the two channels, Ku Band and Ka band, respectively. A basic quality of service using the base layer stream can be provided even when the enhancement layer stream over Ka band channel is unavailable by rain attenuation. A rain fall model is simulated using Barlett-Lewis Pulse model and is used in the operation test of the implemented sacalable satellite broadcasting system.

• JSTS:Journal of Semiconductor Technology and Science
• /
• v.2 no.2
• /
• pp.141-146
• /
• 2002

The effect of the wide-bandgap Schottky barrier enhancement cap layer on the performance of metal-semiconductor-metal photodetectors (MSMPD's) is presented. Judged by the dc characteristics, no considerable increase in recombination loss of carriers is resulted by the incorporation of the cap layer. However, about 45% of the detection efficiency is lost for the cap-layered MSMPD's even with a graded layer incorporated under pulse operation, and it was found to be due mainly to the capturing and slow release of the photocarriers at the heterointerface. The loss mechanism of the pulse detection efficiency is believed to be responsible for the intersymbol interference and the increased bit-error-rate (BER) observed in MSMPD's when used with a high bit rate pseudo-random-bit-stream (PRBS) data pattern.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.9 no.12
• /
• pp.1405-1414
• /
• 2014

The office management service platform configured with office's sensor devices, G/DSCL(Gateway/Device Service Capability Layer), NSCL(Network Service Capability Layer), and NA(Network Application). In this paper, we designed gateway resource tree and service scenario to fit the office management service and demonstrated appropriate operation of the office management service through intelligent functional modeling.

• 한국정보디스플레이학회:학술대회논문집
• /
• 2003.07a
• /
• pp.543-546
• /
• 2003

We studied the dielectric layer effect on the electro-optical (EO) properties of liquid crystal (LC) devices together with numerical simulations. Recently, it has been reported that the surface dielectric layer affects significantly the EO performances of LC microlens arrays and wide-viewing LC displays. it is found that the operation voltage of the LC device decreases with increasing the dielectric constant or with decreasing the thickness of the dielectric polymer layer. The experimental data agree well with theoretical results predicted in a simple dielectric model within the continuum formalism.