• Journal of Broadcast Engineering
• /
• v.16 no.6
• /
• pp.902-915
• /
• 2011

Recently in the area-based stereo matching field, Adaptive Support-Weight (ASW) method that weights matching cost adaptively according to the luminance intensity and the geometric difference shows promising matching performance. However, ASW requires more computational cost than other matching algorithms do and its real-time implementation becomes impractical. By applying Integral Histogram technique after approximating to the Bilateral filter equation, the computational time of ASW can be restricted in constant time regardless of the support window size. However, Integral Histogram technique causes loss of the matching accuracy during approximation process of the original ASW equation. In this paper, we propose a novel algorithm that maintains the ASW algorithm's matching accuracy while reducing the computational costs. In the proposed algorithm, we propose Sub-Block method that groups the pixels within the support area. We also propose the method adjusting the disparity search range depending on edge information. The proposed technique reduces the calculation time efficiently while improving the matching accuracy.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.12 no.3
• /
• pp.231-238
• /
• 2002

As a modeling method where the merits of fuzzy inference system and evolutionary computation are put together, evolutionary fuzzy modeling performs global approximate optimization. The paper proposes fuzzy clustering as fuzzy rule generation process which is one of the most important steps in evolutionary fuzzy modeling. With application of fuzzy clustering into the experiment or simulation results, fuzzy rules which properly describe non-linear and complex design problem can be obtained. The efficiency of evolutionary fuzzy modeling can be improved utilizing the membership degrees of data to clusters from the results of fuzzy clustering. To ensure the validity of the proposed method, the real design problem of an automotive inner trim is applied and the global approximation is achieved. Evolutionary fuzzy modeling is performed for several cases which differ in the number of clusters and the criterion of rule selection and their results are compared to prove that the proposed method can provide proper fuzzy rules for a given system and reduce computation time while maintaining the errors of modeling as a satisfactory level.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.44 no.3
• /
• pp.23-32
• /
• 2007

The H.264/AVC video coding standard uses rate distortion optimization (RDO) method to improve the compression performance in the intra prediction. The complexity and computational load are increased more than previous standard by using this method, even though this standard selects the best coding mode for the current macroblock. This paper proposes a fast intra mode decision algorithm for H.264/AVC encoder based on dominant edge direction (DED). To apply the idea, this algorithm uses the approximation of discrete cosine transform (DCT) coefficient. By detecting the DED, 3 modes instead of 9 modes are chosen for RDO calculation to decide the best mode in the $4{\times}4$ luma block. As for the $16{\times}16$ luma and $8{\times}8$ chroma block, instead of 4 modes, only 2 modes are searched. Experimental results show that the computation time of the proposed algorithm is decreased to about 72% of the full search method with negligible quality loss.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.13 no.8
• /
• pp.1593-1600
• /
• 2009

There are two major algorithms to find a square root of floating point number, one is the Newton_Raphson algorithm and GoldSchmidt algorithm which calculate it approximately by iterating multiplications and the other is SRT algorithm which calculates it exactly by iterating subtractions. This paper proposes an exact floating point square root algorithm using only multiplication. At first an approximate inverse square root is calculated by Newton_Raphson algorithm, and then an exact square root algorithm by reducing an error in it and a compensation algorithm of it are proposed. The proposed algorithm is verified to calculate all of numbers in a single precision floating point number and 1 billion random numbers in a double precision floating point number. The proposed algorithm requires only the multipliers without another hardware, so it can be widely used in an embedded system and mobile production which requires an efact square root of floating point number.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.13 no.1
• /
• pp.47-57
• /
• 2003

In this paper, we firstly evaluate the resistance of the reduced 5-round version of the block cipher CIKS-1 against linear cryptanalysis(LC) and show that we can attack full-round CIKS-1 with \ulcorner56-bit key through the canonical extension of our attack. A feature of the CIKS-1 is the use of both Data-Dependent permutations(DDP) and internal key scheduling which consist in data dependent transformation of the round subkeys. Taking into accout the structure of CIKS-1 we investigate linear approximation. That is, we consider 16 linear approximations with p=3/4 for 16 parallel modulo $2^2$ additions to construct one-round linear approximation and derive one-round linear approximation with the probability P=1/2+$2^{-17}$ by Piling-up lemma. Then we present 3-round linear approximation with 1/2+$2^{-17}$ using this one-round approximation and attack the reduced 5-round CIKS-1 with 64-bit block by LC. In conclusion we present that our attack requires $2^{38}$chosen plaintexts with a probability of success of 99.9% and about $2^{67-7}$encryption times to recover the last round key.(But, for the full-round CIKS-1, our attack requires about $2^{166}$encryption times)

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.13 no.4
• /
• pp.727-736
• /
• 2009

In this paper, the $arccos(cos^{-1})$ arithmetic unit for mobile graphics accelerator is designed. The mobile vector graphics applications need tight area, execution time, power dissipation, and accuracy constraints compared to desktop PC applications. The designed processor adopts 2nd-order polynomial approximation scheme based on IEEE floating point data format to satisfy speed and accuracy conditions and reduces area via hardware sharing structure. The arccosine processor consists of 15,280 gates and its estimated operating frequency is about 125Mhz at operating condition of $0.35{\mu}m$ CMOS technology. Because the processor can execute arccosine function within 7 clock cycles, it has about 17 MOPS(million arccos operations per second) execution rate and can be applicable to mobile OpenVG processor. And because of its flexible architecture, it can be applicable to the various transcendental functions such as exponential, trigonometric and logarithmic functions via replacement of ROM and minor hardware modification.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.44 no.5
• /
• pp.58-65
• /
• 2007

As multimedia portable devices become popular, the amount of computation for processing data including video compression has significantly increased. Various researches for low power consumption of the mobile devices and real time processing have been reported. Motion Estimation is responsible for 67% of H.264 encoder complexity. In this research, a new circuit is designed for motion estimation. The new circuit uses motion prediction based on approximate SAD, Alternative Row Scan (ARS), DAU, and FDVS algorithms. Our new method can reduce the amount of computation by 75% when compared to multi-frame motion estimation suggested in JM8.2. Furthermore, optimal number and size of reference frame blocks are determined to reduce computation without affecting the PSNR. The proposed Motion Estimation method has been verified by using the hardware and software Co-Simulation with iPROVE. It can process 30 CIF frames/sec at 50MHz.

• Journal of Broadcast Engineering
• /
• v.17 no.1
• /
• pp.195-198
• /
• 2012

The N basis functions are typically chosen so that Surface reflectance functions(SRFs) and spectral power distributions (SPDs) can be accurately reconstructed from their N-dimensional vector codes. Typical rendering applications assume that the resulting mapping is an isomorphism where vector operations of addition, scalar multiplication, component-wise multiplication on the N-vectors can be used to model physical operations such as superposition of lights, light-surface interactions and inter-reflection. The vector operations do not mirror the physical. However, if the choice of basis functions is restricted to characteristic functions then the resulting map between SPDs/SRFs and N-vectors is anisomorphism that preserves the physical operations needed in rendering. This paper will show how to select optimal characteristic function bases of any dimension N (number of basis functions) and also evaluate how accurately a large set of Munsell color chips can approximated as basis functions of dimension N.

• The Journal of the Acoustical Society of Korea
• /
• v.39 no.6
• /
• pp.505-514
• /
• 2020

Since a large amount of training data is typically needed to train Deep Neural Networks (DNNs), a parallel training approach is required to train the DNNs. The Stochastic Gradient Descent (SGD) algorithm is one of the most widely used methods to train the DNNs. However, since the SGD is an inherently sequential process, it requires some sort of approximation schemes to parallelize the SGD algorithm. In this paper, we review various efforts on parallelizing the SGD algorithm, and analyze the computational overhead, communication overhead, and the effects of the approximations.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.10 no.12
• /
• pp.3702-3707
• /
• 2009

In this paper, a new approximated MAP algorithm for soft bit decision from QAM symbols is proposed for Gray Coded QAM signals, based on the Max-Log-MAP and a Gray coded QAM signal can be separated into independent two Gray coded PAM signal, M-PAM on I axis with M symbols and N-PAM on Q axis with N symbols. The Max-Log-MAP used distance comparisons between symbols to get the soft bit decision instead of mathematical exponential or logarithm functions. But in accordance with the increase of the number of symbols, the number of comparisons also increase with high complexity. The proposed algorithm is used with the Euclidean distance and constituted with plain arithmetic functions, thus we can know intuitively that the algorithm has low implementing complexity comparing to conventional ones.