• Journal of Broadcast Engineering
• /
• v.24 no.5
• /
• pp.703-712
• /
• 2019

Single image Super-Resolution (SISR) aims to generate a visually pleasing high-resolution image from its degraded low-resolution measurement. In recent years, deep learning - based super - resolution methods have been actively researched and have shown more reliable and high performance. A typical method is WaveletSRNet, which restores high-resolution images through wavelet coefficient learning based on feature maps of images. However, there are two disadvantages in WaveletSRNet. One is a big processing time due to the complexity of the algorithm. The other is not to utilize feature maps efficiently when extracting input image's features. To improve this problems, we propose an efficient single image super resolution method, named RDB-WaveletSRNet. The proposed method uses the residual dense block to effectively extract low-resolution feature maps to improve single image super-resolution performance. We also adjust appropriated growth rates to solve complex computational problems. In addition, wavelet packet decomposition is used to obtain the wavelet coefficients according to the possibility of large scale ratio. In the experimental result on various images, we have proven that the proposed method has faster processing time and better image quality than the conventional methods. Experimental results have shown that the proposed method has better image quality by increasing 0.1813dB of PSNR and 1.17 times faster than the conventional method.

• Journal of Internet Computing and Services
• /
• v.22 no.2
• /
• pp.59-68
• /
• 2021

As artificial intelligence(AI) technologies, which have made rapid growth recently, began to be applied to the marine environment such as ships, there have been active researches on the application of CNN-based models specialized for digital videos. In E-Navigation service, which is combined with various technologies to detect floating objects of clash risk to reduce human errors and prevent fires inside ships, real-time processing is of huge importance. More functions added, however, mean a need for high-performance processes, which raises prices and poses a cost burden on shipowners. This study thus set out to propose a method capable of processing information at a high rate while maintaining the accuracy by applying Quantization techniques of a deep learning model. First, videos were pre-processed fit for the detection of floating matters in the sea to ensure the efficient transmission of video data to the deep learning entry. Secondly, the quantization technique, one of lightweight techniques for a deep learning model, was applied to reduce the usage rate of memory and increase the processing speed. Finally, the proposed deep learning model to which video pre-processing and quantization were applied was applied to various embedded boards to measure its accuracy and processing speed and test its performance. The proposed method was able to reduce the usage of memory capacity four times and improve the processing speed about four to five times while maintaining the old accuracy of recognition.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.9 no.4
• /
• pp.11-22
• /
• 1999

Least-mean-square(LMS) adaptive filters have proven to be extremely useful in a number of signal processing tasks. However LMS adaptive filter suffer from a slow rate of convergence for a given steady-state mean square error as compared to the behavior of recursive least squares adaptive filter. In this paper an efficient signal interference control technique is introduced to improve the convergence speed of LMS algorithm with tap weighted vectors updating which were controled by reusing data which was abandoned data in the Adaptive transversal filter in the scheme with data recycling buffers. The computer simulation show that the character of convergence and the value of MSE of proposed algorithm are faster and lower than the existing LMS according to increasing the step-size parameter $\mu$ in the experimentally computed. learning curve. Also we find that convergence speed of proposed algorithm is increased by (B+1) time proportional to B which B is the number of recycled data buffer without complexity of computation. Adaptive transversal filter with proposed data recycling buffer algorithm could efficiently reject ISI of channel and increase speed of convergence in avoidance burden of computational complexity in reality when it was experimented having the same condition of LMS algorithm.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.11 no.1
• /
• pp.13-23
• /
• 2001

This paper implemented into hardware SEED which is the KOREA standard 128-bit block cipher. First, at the respect of hardware implementation, we compared and analyzed SEED with AES finalist algorithms - MARS, RC6, RIJNDAEL, SERPENT, TWOFISH, which are secret key block encryption algorithms. The encryption of SEED is faster than MARS, RC6, TWOFISH, but is as five times slow as RIJNDAEL which is the fastest. We propose a SEED hardware architecture which improves the encryption speed. We divided one round into three parts, J1 function block, J2 function block J3 function block including key mixing block, because SEED repeatedly executes the same operation 16 times, then we pipelined one round into three parts, J1 function block, J2 function block, J3 function block including key mixing block, because SEED repeatedly executes the same operation 16 times, then we pipelined it to make it more faster. G-function is implemented more easily by xoring four extended 4 byte SS-boxes. We tested it using ALTERA FPGA with Verilog HDL. If the design is synthesized with 0.5 um Samsung standard cell library, encryption of ECB and decryption of ECB, CBC, CFB, which can be pipelined would take 50 clock cycles to encrypt 384-bit plaintext, and hence we have 745.6 Mbps assuming 97.1 MHz clock frequency. Encryption of CBC, OFB, CFB and decryption of OFB, which cannot be pipelined have 258.9 Mbps under same condition.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.12 no.6
• /
• pp.17-28
• /
• 2002

XTR is a new method to represent elements of a subgroup of a multiplicative group of a finite field GF( $p^{6m}$) and it can be generalized to the field GF( $p^{6m}$)$^{[6,9]}$ This paper progress optimal extention fields for XTR among Galois fields GF ( $p^{6m}$) which can be aplied to XTR. In order to select such fields, we introduce a new notion of Generalized Opitimal Extention Fields(GOEFs) and suggest a condition of prime p, a defining polynomial of GF( $p^{2m}$) and a fast method of multiplication in GF( $p^{2m}$) to achieve fast finite field arithmetic in GF( $p^{2m}$). From our implementation results, GF( $p^{36}$ )longrightarrowGF( $p^{12}$ ) is the most efficient extension fields for XTR and computing Tr( $g^{n}$ ) given Tr(g) in GF( $p^{12}$ ) is on average more than twice faster than that of the XTR system on Pentium III/700MHz which has 32-bit architecture.$^{［6,10］/ ［6,10］/6,10］}$

• Journal of the Institute of Convergence Signal Processing
• /
• v.21 no.3
• /
• pp.148-153
• /
• 2020

In order to obtain high accuracy with an convolutional neural network(CNN), it is necessary to set the optimal hyperparameters. However, the exact value of the hyperparameter that can make high performance is not known, and the optimal hyperparameter value is different based on the type of the dataset, therefore, it is necessary to find it through various experiments. In addition, since the range of hyperparameter values is wide and the number of combinations is large, it is necessary to find the optimal values of the hyperparameters after the experimental design in order to save time and computational costs. In this paper, we suggest an algorithm that use the design of experiments and grid search algorithm to determine the optimal hyperparameters for a classification problem. This algorithm determines the optima values of the hyperparameters that yields high performance using the factorial design of experiments. It is shown that the amount of computational time can be efficiently reduced and the accuracy can be improved by performing a grid search after reducing the search range of each hyperparameter through the experimental design. Moreover, Based on the experimental results, it was shown that the learning rate is the only hyperparameter that has the greatest effect on the performance of the model.

• Journal of Aerospace System Engineering
• /
• v.15 no.1
• /
• pp.80-85
• /
• 2021

In this study, for conducting a static load test of an external fuel tank used for an aircraft, the flight load acting on the external fuel tank was converted to the test load and the suitability of the converted test loads was confirmed. In order to calculate the test load from the flight load, the external fuel tank was divided into several sections. Shear load, moment by unit shear load, and unit moment were calculated for each section. Test loads for each section were then calculated by computing the shear load, the moment of each section, and flight load condition. In actual static load tests, it might not be possible to impose the test load in the calculated position due to physical constraints. Therefore, after determining positions in which the load could be imposed in the actual test, the test load calculated for each section was redistributed to selected positions. Finally, a test load plan was established by applying a whiffle tree to enhance the efficiency of the test performance while making it easier to operate the actuator. The reliability of the test load plan was verified by comparing it with flight load conditions.

• Korean Journal of Remote Sensing
• /
• v.36 no.6_2
• /
• pp.1605-1613
• /
• 2020

The tidal channel is a coastal sedimentary terrain that plays the most important role in the formation and development of tidal flats, and is considered a very important index for understanding and distribution of tidal flat sedimentation/erosion terrain. The purpose of this study is to understand the changes in tidal channels by a period after the opening of the floodgate of the seawall in the reclaimed land of Sihwa Lake using KOMPSAT high-resolution multispectral satellite image data and to evaluate the applicability and efficiency of high-resolution satellite images. KOMPSAT 2 and 3 images were used for extraction of the tidal channels' lineaments in 2009, 2014, and 2019 and were applied to supervised classification method based on Principal Component Analysis (PCA), Artificial Neural Net (ANN), Matched Filtering (MF), and Spectral Angle Mapper (SAM) and band ratio techniques using Normalized Difference Water Index (NDWI) and MF/SAM. For verification, a numerical map of the National Geographic Information Service and Landsat 7 ETM+ image data were utilized. As a result, KOMPSAT data showed great agreement with the verification data compared to the Landsat 7 images for detecting a direction and distribution pattern of the tidal channels. However, it has been confirmed that there will be limitations in identifying the distribution of tidal channels' density and providing meaningful information related to the development of the sedimentary process. This research is expected to present the possibility of utilizing KOMPSAT image-based high-resolution remote exploration as a way of responding to domestic intertidal environmental issues, and to be used as basic research for providing multi-platform-image-based convergent thematic maps and topics.

• Journal of IKEEE
• /
• v.26 no.4
• /
• pp.633-638
• /
• 2022

Rencently, neuromorphic systems of spiking neural networks (SNNs) that imitate the human brain have attracted attention. Neuromorphic technology has the advantage of high speed and low power consumption in cognitive applications and processing. Resistive random-access memory (RRAM) for SNNs are the most efficient structure for parallel calculation and perform the gradual switching operation of spike-timing-dependent plasticity (STDP). RRAM as synaptic device operation has low-power processing and expresses various memory states. However, the integration of RRAM device causes high switching voltage and current, resulting in high power consumption. To reduce the operation voltage of the RRAM, it is important to develop new materials of the switching layer and metal electrode. This study suggested a optimized new structure that is the Metal/Al₂O₃/HfO_x/SWCNTs/N+silicon (MOCS) with single-walled carbon nanotubes (SWCNTs), which have excellent electrical and mechanical properties in order to lower the switching voltage. Therefore, we show an improvement in the gradual switching behavior and low-power I/V curve of SWCNTs-based memristors.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.10
• /
• pp.29-41
• /
• 2022

Recently, with the development of deep learning technology, a variety of huge models with excellent performance have been devised by pre-training massive amounts of text data. However, in order for such a model to be applied to real-life services, the inference speed must be fast and the amount of computation must be low, so the technology for model compression is attracting attention. Knowledge distillation, a representative model compression, is attracting attention as it can be used in a variety of ways as a method of transferring the knowledge already learned by the teacher model to a relatively small-sized student model. However, knowledge distillation has a limitation in that it is difficult to solve problems with low similarity to previously learned data because only knowledge necessary for solving a given problem is learned in a teacher model and knowledge distillation to a student model is performed from the same point of view. Therefore, we propose a heterogeneous knowledge distillation method in which the teacher model learns a higher-level concept rather than the knowledge required for the task that the student model needs to solve, and the teacher model distills this knowledge to the student model. In addition, through classification experiments on about 18,000 documents, we confirmed that the heterogeneous knowledge distillation method showed superior performance in all aspects of learning efficiency and accuracy compared to the traditional knowledge distillation.