• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.4
• /
• pp.101-110
• /
• 1996

Color image quantization is a process of selecting a set of colors to display an image with some representative colors without noticeable perceived difference. It is very important in many applications to display a true color image in a low cost color monitor or printer. the basic problem is how to display 224 colors with 256 or less colors, called color palette. In this paper, we propose an algorithm to design the 256 or less size color palette by using spatial maskin geffect of HVS and subjective distortion measure weighted by color palette by using spatial masking effect of HVS and subjective distortion measure weighted by color activity in 4*4 local region in any color image. The proposed algorithm consists of octal prequantization and subdivision quantization processing step using the distortion measure and modified Otsu's between class variance maximization method. The experimental results show that the proposed algorithm has higher visual quality and needs less consuming time than conventional algorithms.

• Phonetics and Speech Sciences
• /
• v.1 no.2
• /
• pp.3-8
• /
• 2009

In an effort to improve speech recognition, we investigated the effect of the number of clusters. In usual LBG clustering, the number of codebook clusters is doubled on each bifurcation and hence cannot be chosen arbitrarily in a natural way. To have the number of clusters at our control, we combined adaptive resonance theory (ART2) with LBG and perform the clustering in two stages. The codebook thus formed was used in subsequent processing of fuzzy vector quantization (FVQ) and HMM for speech recognition tests. Compared to conventional LBG, our method was shown to reduce the best recognition error rate by 0${\sim$}0.9% depending on the vocabulary size. The result also showed that between 400 and 800 would be the optimal number of clusters in the limit of small and large vocabulary speech recognitions of isolated words, respectively.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.31B no.8
• /
• pp.99-108
• /
• 1994

In this paper an image compression technique based on fractal theory using iterated contractive transformation is analysed and an improved image coder is suggested. Existing methods used the classifier proposed by Ramamurthi and Gersho which utilize the properties of neighboring pixels in the spatial domain. In this paper DCT-based classification is applied to 512$\times$512 images and PSNR improvement of 0.4~2.7 dB is obtained at lower bit rate over conventional algorithms. In addition the effect of varying the domain block size and quantization step size of the luminance shift parameter on the compression ratio and the image quality is compared and analysed.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.30 no.12C
• /
• pp.1231-1238
• /
• 2005

This paper presents a new fast encoding algorithm for image vector quantization that incorporates the partial distances of multiple features with a multidimensional look-up table (LUT). Although the methods which were proposed earlier use the multiple features, they handles the multiple features step by step in terms of searching order and calculating process. On the other hand, the proposed algorithm utilizes these features simultaneously with the LUT. This paper completely describes how to build the LUT with considering the boundary effect for feasible memory cost and how to terminate the current search by utilizing partial distances of the LUT Simulation results confirm the effectiveness of the proposed algorithm. When the codebook size is 256, the computational complexity of the proposed algorithm can be reduced by up to the $70\%$ of the operations required by the recently proposed alternatives such as the ordered Hadamard transform partial distance search (OHTPDS), the modified $L_2-norm$ pyramid ($M-L_2NP$), etc. With feasible preprocessing time and memory cost, the proposed algorithm reduces the computational complexity to below the $2.2\%$ of those required for the exhaustive full search (EFS) algorithm while preserving the same encoding quality as that of the EFS algorithm.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.23 no.9
• /
• pp.814-822
• /
• 2013

For the fault diagnosis of a mechanical system, pattern recognition methods have being used frequently in recent research. Hidden Markov model(HMM) and artificial neural network(ANN) are typical examples of pattern recognition methods employed for the fault diagnosis of a mechanical system. In this paper, a hybrid method that combines HMM and ANN for the fault diagnosis of a mechanical system is introduced. A rotating blade which is used for a wind turbine is employed for the fault diagnosis. Using the HMM/ANN hybrid model along with the numerical model of the rotating blade, the location and depth of a crack as well as its presence are identified. Also the effect of signal to noise ratio, crack location and crack size on the success rate of the identification is investigated.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.21 no.4
• /
• pp.822-833
• /
• 1996

The video coding for very low bit rate has recently received considerable attention, but conventional block based transform coding schemes suffer from the blocking effect for the constraints of bit rates. In this paper, we present a video coding sysem suing multi-resolution motion estimation/compensation with variable size block(VMRME/C) and multi-resolution vector quantization(MRVQ) in wavelet transform domain for very low bit rate coding. It is shown that the presented scheme has better performance in the peak signal-to-nose ratio(RSNR) by 0.2-0.6 dB as well as subjective quality than that of conventional block based transform video coding techniques(especially, H. 263 which is DCT based video coding).

• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.05a
• /
• pp.373-376
• /
• 2021

Deep neural networks have achieved almost human-level results in various tasks and have become popular in the broad artificial intelligence domains. Uncertainty estimation is an on-demand task caused by the black-box point estimation behavior of deep learning. The deep ensemble provides increased accuracy and estimated uncertainty; however, linearly increasing the size makes the deep ensemble unfeasible for memory-intensive tasks. To address this problem, we used model pruning and quantization with a deep ensemble and analyzed the effect in the context of uncertainty metrics. We empirically showed that the ensemble members' disagreement increases with pruning, making models sparser by zeroing irrelevant parameters. Increased disagreement implies increased uncertainty, which helps in making more robust predictions. Accordingly, an energy-efficient compressed deep ensemble is appropriate for memory-intensive and uncertainty-aware tasks.

• Journal of Electrical Engineering and Technology
• /
• v.13 no.5
• /
• pp.2074-2079
• /
• 2018

Global state space's optimal policy is used for offline controller in the form of table by using Dynamic Programming. If an optimal policy table has a large amount of control data, it is difficult to use the system in a low capacity system. To resolve these problem, controller using the compressed optimal policy table is proposed in this paper. A DCT is used for compression method and the cosine function is used as a basis. The size of cosine function decreased as the frequency increased. In other words, an essential information which is used for restoration is concentrated in the low frequency band and a value of small size that belong to a high frequency band could be discarded by quantization because high frequency's information doesn't have a big effect on restoration. Therefore, memory could be largely reduced by removing the information. The compressed output is stored in memory of embedded system in offline and optimal control input which correspond to state of plant is computed by interpolation with Inverse DCT in online. To verify the performance of the proposed controller, computer simulation was accomplished with a one link inverted pendulum.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2012.02a
• /
• pp.213-213
• /
• 2012

It has been known that quantum confinement effect of CdSe nanocrystal was observed by increasing the number of deposition cycle using successive ionic layer adsorption and reaction (SILAR) method. Here, we report on thermally-induced quantum confinement effect of CdSe at the given cycle number using spin-coating technology. A cation precursor solution containing $0.3\;M\;Cd(NO_3)_2{\cdot}4H_2O$ is spun onto a $TiO_2$ nanoparticulate film, which is followed by spinning an anion precursor solution containing $0.3\;M\;Na_2\;SeSO_3$ to complete one cycle. The cycle is repeated up to 10 cycles, where the spin-coated $TiO_2$ film at each cycle is heated at temperature ranging from $100^{\circ}C$ to $250^{\circ}C$. The CdSe-sensitized $TiO_2$ nanostructured film is contacted with polysulfide redox electrolyte to construct photoelectrochemical solar cell. Photovoltaic performance is significantly dependent on the heat-treatment temperature. Incident photon-to-current conversion efficiency (IPCE) increases with increasing temperature, where the onset of the absorption increases from 600 nm for the $100^{\circ}C$- to 700 nm for the $150^{\circ}C$- and to 800 nm for the $200^{\circ}C$- and the $250^{\circ}C$-heat treatment. This is an indicative of quantum size effect. According to Tauc plot, the band gap energy decreases from 2.09 eV to 1.93 eV and to 1.76 eV as the temperature increases from $100^{\circ}C$ to $150^{\circ}C$ and to $200^{\circ}C$ (also $250^{\circ}C$), respectively. In addition, the size of CdSe increases gradually from 4.4 nm to 12.8 nm as the temperature increases from $100^{\circ}C$ to $250^{\circ}C$. From the differential thermogravimetric analysis, the increased size in CdSe by increasing the temperature at the same deposition condition is found to be attributed to the increase in energy for crystallization with $dH=240cal/^{\circ}C$. Due to the thermally induced quantum confinement effect, the conversion efficiency is substantially improved from 0.48% to 1.8% with increasing the heat-treatment temperature from $100^{\circ}C$ to $200^{\circ}C$.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.19 no.7
• /
• pp.1203-1212
• /
• 1994

This paper proposes a variable block size vector quantization based on a biorthogonal wavelet transform for image compression. An image is first decomposed with the biorthogonal wavelet transform into multiresolution image and the wavelet coefficients of the middle frequency bands are segmented using the quadtree sturcture to extract the perceptually important regions in the middle frequency bands. A sedges of middle frequency bands exist the corresponding position of high frequency bands, the complicated quadtree structure of middle frequency bands is equally applied to the high frequency bands. Therefore the overhaed information of the quadtree codes needed to segment the high frequency bands can be reduced. The segmented subblocks are encoded with the codebook designed at the each scales and directions. The simulation results showed that the proposed methods could reproduce higher quality image with bit rate reduced about 20(%) than of the preceding VQ method and sufficiently reduce the bolck effect and the edge degradation.