• Journal of Internet Computing and Services
• /
• v.22 no.6
• /
• pp.141-150
• /
• 2021

Based on Artificial Intelligence technology, AI-enabled warfare is expected to become the main issue in the future warfare. Natural language processing technology is a core technology of AI technology, and it can significantly contribute to reducing the information burden of underrstanidng reports, information objects and intelligences written in natural language by commanders and staff. In this paper, we propose a Language model-based Multi-source Information Integration (LAMII) framework to reduce the information overload of commanders and support rapid decision-making. The proposed LAMII framework consists of the key steps of representation learning based on language models in self-supervsied way and document integration using autoencoders. In the first step, representation learning that can identify the similar relationship between two heterogeneous sentences is performed using the self-supervised learning technique. In the second step, using the learned model, documents that implies similar contents or topics from multiple sources are found and integrated. At this time, the autoencoder is used to measure the information redundancy of the sentences in order to remove the duplicate sentences. In order to prove the superiority of this paper, we conducted comparison experiments using the language models and the benchmark sets used to evaluate their performance. As a result of the experiment, it was demonstrated that the proposed LAMII framework can effectively predict the similar relationship between heterogeneous sentence compared to other language models.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.45 no.2
• /
• pp.1-9
• /
• 2008

As the security requirements of mobile phones have been increasing, there have been extensive researches using one biometric feature (e.g., an iris, a fingerprint, or a face image) for authentication. Due to the limitation of uni-modal biometrics, we propose a method that combines face and iris images in order to improve accuracy in mobile environments. This paper presents four advantages and contributions over previous research. First, in order to capture both face and iris image at fast speed and simultaneously, we use a built-in conventional mega pixel camera in mobile phone, which is revised to capture the NIR (Near-InfraRed) face and iris image. Second, in order to increase the authentication accuracy of face and iris, we propose a score level fusion method based on SVM (Support Vector Machine). Third, to reduce the classification complexities of SVM and intra-variation of face and iris data, we normalize the input face and iris data, respectively. For face, a NIR illuminator and NIR passing filter on camera are used to reduce the illumination variance caused by environmental visible lighting and the consequent saturated region in face by the NIR illuminator is normalized by low processing logarithmic algorithm considering mobile phone. For iris, image transform into polar coordinate and iris code shifting are used for obtaining robust identification accuracy irrespective of image capturing condition. Fourth, to increase the processing speed on mobile phone, we use integer based face and iris authentication algorithms. Experimental results were tested with face and iris images by mega-pixel camera of mobile phone. It showed that the authentication accuracy using SVM was better than those of uni-modal (face or iris), SUM, MAX, NIN and weighted SUM rules.

• Journal of Convergence for Information Technology
• /
• v.10 no.10
• /
• pp.1-8
• /
• 2020

The need for storage, management, and retrieval techniques for alternative data has emerged as technologies based on data generated from business activities conducted by enterprises have emerged as the key to business success in recent years. Existing big data platform systems must load a large amount of data generated in real time without delay to process unstructured data, which is an alternative data, and efficiently manage storage space by utilizing a deduplication system of different storages when redundant data occurs. In this paper, we propose a multi-layer distributed data deduplication process system using the similarity of the Cuckoo hashing filter technique considering the characteristics of big data. Similarity between virtual machines is applied as Cuckoo hash, individual storage nodes can improve performance with deduplication efficiency, and multi-layer Cuckoo filter is applied to reduce processing time. Experimental results show that the proposed method shortens the processing time by 8.9% and increases the deduplication rate by 10.3%.

• The KIPS Transactions:PartB
• /
• v.13B no.2 s.105
• /
• pp.105-114
• /
• 2006

In ubiquitous environment where various kinds of computers are embedded in persons, objects and environment and they are interconnected and can be used in my place as necessary, different types of data need to be exchanged between heterogeneous machines through home network. In the environment, the efficient processing, transmission and monitoring of image data are essential technologies. We need to make research not only on traditional image processing such as spatial and visual resolution, color expression and methods of measuring image quality but also on transmission rate on home network that has a limited bandwidth. The present study proposes a new motion vector estimation algorithm for transmitting, processing and controlling image data, which is the core part of contents in home network situation and, using algorithm, implements a real time monitoring system of multi dimensional images transmitted from multiple cameras. Image data of stereo cameras to be transmitted in different environment in angle, distance, etc. are preprocessed through reduction, magnification, shift or correction, and compressed and sent using the proposed metamorphosis hierarchical motion vector estimation algorithm for the correction of motion. The proposed algorithm adopts advantages and complements disadvantages of existing motion vector estimation algorithms such as whole range search, three stage search and hierarchical search, and estimates efficiently the motion of images with high variation of brightness using an atypical small size macro block. The proposed metamorphosis hierarchical motion vector estimation algorithm and implemented image systems can be utilized in various ways in ubiquitous environment.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.10 no.3
• /
• pp.503-510
• /
• 2006

Readout circuit is to convert signal occurred in a defector into suitable signal for image signal processing. In general, it has to possess functions of impedance matching with perception element, amplification, noise reduction and cell selection. It also should satisfies conditions of low-power, low-noise, linearity, uniformity, dynamic range, excellent frequency-response characteristic, and so on. The technical issues in developing image processing equipment for focal plane way (FPA) can be categorized as follow: First, ultraviolet (UV) my detector material and fine processing technology. Second, ReadOut IC (ROIC) design technology to process electric signal from detector. Last, package technology for hybrid bonding between detector and ROIC. ROIC enables intelligence and multi-function of image equipment. It is a core component for high value added commercialization ultimately. Especially, in development of high-resolution image equipment ROIC, it is necessary that high-integrated and low-power circuit design technology satisfied with design specifications such as detector characteristic, signal dynamic range, readout rate, noise characteristic, ceil pitch, power consumption and so on. In this paper, we implemented a $8\times8$ FPA prototype ROIC for reduction of period and cost. We tested unit block and overall functions of designed $8\times8$ FPA ROIC. Also, we manufactured ROIC control and image boards, and then were able to verify operation of ROIC by confirming detected image from PC's monitor through UART(Universal Asynchronous Receiver Transmitter) communication.

• JSTS:Journal of Semiconductor Technology and Science
• /
• v.17 no.3
• /
• pp.446-457
• /
• 2017

In order to reduce the size of frame memory or bus bandwidth, frame memory compression (FMC) recompresses reconstructed or reference frames of video codecs. This paper proposes a novel FMC design based on discrete wavelet transform (DWT) - set partitioning in hierarchical trees (SPIHT), which supports fine-scalable throughput and is area-efficient. In the proposed design, multi-cores with small block sizes are used in parallel instead of a single core with a large block size. In addition, an appropriate pipelining schedule is proposed. Compared to the previous design, the proposed design achieves the processing speed which is closer to the target system speed, and therefore it is more efficient in hardware utilization. In addition, a scheme in which two passes of SPIHT are merged into one pass called merged refinement pass (MRP) is proposed. As the number of shifters decreases and the bit-width of remained shifters is reduced, the size of SPIHT hardware significantly decreases. The proposed FMC encoder and decoder designs achieve the throughputs of 4,448 and 4,000 Mpixels/s, respectively, and their gate counts are 76.5K and 107.8K. When the proposed design is applied to high efficiency video codec (HEVC), it achieves 1.96% lower average BDBR and 0.05 dB higher average BDPSNR than the previous FMC design.

• Transactions of Materials Processing
• /
• v.28 no.6
• /
• pp.361-367
• /
• 2019

Microstructure evolution and tensile properties of Al-8mass%Mg alloy casting billet by biaxial alternative forging were investigated in this study. An alternative forging system tailored in this study was used to allow continuous strain accumulations on the alloy workpiece. A finite element (FE) simulation results revealed that the strain was mainly concentrated in the core and that the shear bands developed into a form with an X shape in the cross-section of workpiece after the alternative forging using octangular rod shaped dies. With increasing the forging passes, it was observed that the Al-8mass%Mg alloy workpieces were significantly deformed, and cracks began to form and propagate on the both ends of the forged workpieces after five passes at room temperature. In as-forged microstructures taken by microscopes, twins, clustering of dislocations, and fine subgrains were found. Tensile strengths of the forged specimens showed significant increases depending on the number of forging passes, and a trade-off relationship was observed between the elongation and strength. At room temperature and 100℃ the workpieces showed similar behaviors in microstructural evolution and tensile properties depending on forging passes, while the increase range in strength was reduced at 200℃.

• Computers and Concrete
• /
• v.12 no.2
• /
• pp.211-228
• /
• 2013

The structure analyzed in this paper has particular building style and special structural system. It is a rigid-connected twin-tower skyscraper with asymmetrical distribution of stiffness and masses in two towers. Because of the different stiffness between the north and the south towers, the torsion seismic vibration is significant. In this paper, in order to study the seismic response of the structure under both frequent low-intensity earthquakes as well as rare earthquakes at the levels of intensity 7, the analysis model is built and analyzed with NosaCAD. NosaCAD is an nonlinear structure analysis software based on second-development of AutoCAD with ObjectARX. It has convenient modeling function, high computational efficiency and diversity post-processing functions. The deformations, forces and damages of the structure are investigated based on the analysis. According to the analysis, there is no damage on the structure under frequent earthquakes, and the structure has sufficient capacity and ductility to resist rare earthquakes. Therefore the structure can reach the goal of no damage under frequent earthquakes and no collapse under rare earthquakes. The deformation of the structure is below the limit in Chinese code. The time sequence and distribution of damages on tubes are reasonable, which can dissipate some dynamic energy. At last, according to forces, load-carrying capacity and damage of elements, there are some suggestions on increasing the reinforcement in the core tube at base and in stiffened stories.

• KIPS Transactions on Computer and Communication Systems
• /
• v.4 no.1
• /
• pp.11-14
• /
• 2015

WSN (Wireless Sensor Network) based on low-power is one of the core technologies in the ubiquitous society. In this paper, we present a border surveillance and control system in WSN environment. The system consists of static sensor node, mobile sensor node, static gateway, mobile gateway, server and mobile application. Mobile applications are divided into user mode and manager mode. So users monitor border surveillance through mobile phone and get information of border network environment without time and space constraints. In manager mode, for the flexible operation of nodes, manager can update to the software remotely and adjust the position of the mobile node. And also we implement a suitable multi-hop routing protocol for scalable low-power sensor nodes and confirm that the system operates well in WSN environment.

• Journal of Biomedical Engineering Research
• /
• v.41 no.3
• /
• pp.128-137
• /
• 2020

Cardiac electrophysiology studies often use simulation to predict how cardiac will behave under various conditions. To observe the cardiac tissue movement, it needs to use the high--resolution heart mesh with a sophisticated and large number of nodes. The higher resolution mesh is, the more computation time is needed. To improve computation speed and performance, parallel processing using multi-core processes and network computing resources is performed. In this study, we compared the computational speeds of CPU parallelization and GPU parallelization in virtual heart simulation for efficiently calculating a series of ordinary differential equations (ODE) and partial differential equations (PDE) and determined the optimal CPU and GPU parallelization architecture. We used 2D tissue model and 3D ventricular model to compared the computation performance. Then, we measured the time required to the calculation of ODEs and PDEs, respectively. In conclusion, for the most efficient computation, using GPU parallelization rather than CPU parallelization can improve performance by 4.3 times and 2.3 times in calculations of ODEs and PDE, respectively. In CPU parallelization, it is best to use the number of processors just before the communication cost between each processor is incurred.