• The Journal of the Acoustical Society of Korea
• /
• v.21 no.2
• /
• pp.199-210
• /
• 2002

In this paper, we carried out the study on speech recognition using the KM-Net topology design algorithm based on decision tree state-clustering to improve the performance of acoustic models in speech recognition. The Korean has many allophonic and grammatical rules compared to other languages, so we investigate the allophonic variations, which defined the Korean phonetics, and construct the phoneme question set for phonetic decision tree. The basic idea of the HM-Net topology design algorithm is that it has the basic structure of SSS (Successive State Splitting) algorithm and split again the states of the context-dependent acoustic models pre-constructed. That is, it have generated. the phonetic decision tree using the phoneme question sets each the state of models, and have iteratively trained the state sequence of the context-dependent acoustic models using the PDT-SSS (Phonetic Decision Tree-based SSS) algorithm. To verify the effectiveness of the above algorithm we carried out the speech recognition experiments for 452 words of center for Korean language Engineering (KLE452) and 200 sentences of air flight reservation task (YNU200). Experimental results show that the recognition accuracy has progressively improved according to the number of states variations after perform the splitting of states in the phoneme, word and continuous speech recognition experiments respectively. Through the experiments, we have got the average 71.5%, 99.2% of the phoneme, word recognition accuracy when the state number is 2,000, respectively and the average 91.6% of the continuous speech recognition accuracy when the state number is 800. Also we haute carried out the word recognition experiments using the HTK (HMM Too1kit) which is performed the state tying, compared to share the parameters of the HM-Net topology design algorithm. In word recognition experiments, the HM-Net topology design algorithm has an average of 4.0% higher recognition accuracy than the context-dependent acoustic models generated by the HTK implying the effectiveness of it.

• Journal of Astronomy and Space Sciences
• /
• v.26 no.4
• /
• pp.425-438
• /
• 2009

We reevaluate the Burton equation (Burton et al. 1975) of predicting Dst index using high quality hourly solar wind data supplied by the ACE satellite for the period from 1998 to 2006. Sixty magnetic storms with monotonously decreasing main phase are selected. In order to determine the injection term (Q) and the decay time ($\tau$) of the equation, we examine the relationships between $Dst^*$ and $VS_s$, ${\Delta}Dst^*$ and $VS_s$, and ${\Delta}Dst^*$ and $Dst^*$ during the magnetic storms. For this analysis, we take into account one hour of the propagation time from the ACE satellite to the magnetopause, and a half hour of the response time of the magnetosphere/ring current to he solar wind forcing. The injection term is found to be $Q(nT/h)\;=\;-3.56VS_s$ for $VS_s$ > 0.5mV/m and Q(nT=h) = 0 for $VB_s\;{\leq}\;0.5mV/m$. The $\tau$ (hour) is estimated as $0.060Dst^*\;+\;16.65$ for $Dst^*$ > -175nT and 6.15 hours for $Dst^*\;{\leq}\;-175nT$. Based on these empirical relationships, we predict the 60 magnetic storms and find that the correlation coefficient between the observed and predicted $Dst^*$ is 0.88. To evaluate the performance of our prediction scheme, the 60 magnetic storms are predicted again using the models by Burton et al. (1975) and O'Brien & McPherron (2000a). The correlation coefficients thus obtained are 0.85, the same value for both of the two models. In this respect, our model is slightly improved over the other two models as far as the correlation coefficients is concerned. Particularly our model does a better job than the other two models in predicting intense magnetic storms ($Dst^*\;{< \atop \sim}\;-200nT$).

• Journal of Korean Society of Environmental Engineers
• /
• v.27 no.6
• /
• pp.590-598
• /
• 2005

An adequate method to identify chromium separation, Cr(III) and Cr(VI), in water samples were studied by using High Performance Liquid Chromatography(HPLC) coupled with Inductively Coupled Plasma Mass Spectometer(ICP-MS) equipped with Dynamic Reaction Cell(DRC). The characteristic distribution of Cr(III) and Cr(VI) in the raw water taken at the six water intake stations in Seoul, was analyzed by the method developed by the authors. The chromium species separated by HPLC was isocratically conducted by using tetrabutylammonium phosphate monobasic(1.0 mM TBAP), ethylenediaminetetraacetic acid(0.6 mM EDTA) and 2% v/v methanol as the mobile phase. 5% v/v methanol was used as flushing solvent. A reactive ammonia($NH_3$) gas was used to eliminate the potential interference of $ArC^+$. Several Parameters such as solvent ratio, pH, flow rate and sample injection volume were optimized for the successful separation and reproducibility. Although it has been reported thai the separation sensitivity of Cr(III) is superior to that of Cr(VI), the authors observed Cr(VI) was more sensitive than Cr(III) when ammonia($NH_3$) gas was used as the reaction gas. It took less than 3 minutes to analyze chromium species with this method and the estimated detection limits were $0.061\;{\mu}g/L$ for Cr(III) and $0.052\;{\mu}g/L$, for Cr(VI). According to the results from the analysis on chromium species in the raw water of the six intake stations, the concentrations of Cr(III) ranged from 0.048 to $0.064\;{\mu}g/L$(ave. $0.054\;{\mu}g/L$) while that of Cr(VI) ranged from 0.014 to $0.023\;{\mu}g/L$(ave. $0.019\;{\mu}g/L$). Recovery ratio was very high($90.1{\sim}94.1%$). There were two or three times more Cr(III) than Cr(VI) in the raw water.

• Journal of Korean Society of Forest Science
• /
• v.79 no.1
• /
• pp.56-70
• /
• 1990

Eighteen Korean white pine (P. koraiensis S. et Z.) families were tested in 3 different regions from age 5 to 9. Family and site were significant sources of variation for seedling survival and field growth, whereas the effects of family x site interaction ware relatively small as compared with the former sources of variation. Variance components estimated from the separate and combined sites indicated that the most variabilities were associated with individual trees within plot. Family ${\times}$ site interaction components as a percentage of family variance decreased sharply with age. Heritability estimates varied with testing site and tree age. Combined analyses, however, showed a moderate change in heritability with increasing tree ages, and demonstrated high and stable trends of estimates, particularly in family heritabilities of tree height ($h_F{^2}=0.789-0.798$). The gains estimated from combined analysis have expected maximum or near-maximum efficiencies at age 6 or 7. Given equal intensity of selection, mass selection showed the most efficient gains within and across the sites. However, for the differences between mass and combined selections are small, selection made on the combination of family and within-family would be more effective in improving genetic gains. Indirect selection method indicated that 5-and 6-years height were all good predictors of 9-year-old height with little loss of relative efficiency (less than 10%) as compared with direct family selection at age 9. Phenotypic and genetic correlations computed on the basis of family mean values of height and diameter have shown predominantly high, positive, and statistically significant (1% level) relationships between all tested pairs of traits, which indicates that family growth maintained statistically consistent trends with age. The best families are those that maintained a stable superiority overall sites and ages in growth performance, therefore, it can be suggested that early identification of superior families at age 9 is feasible at age 5 or 6 in Pinus koraiensis S. et Z.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2010.06a
• /
• pp.375-375
• /
• 2010

CdTe as an absorber material is widely used in thin film solar cells with the heterostructure due to its almost ideal band gap energy of 1.45 eV, high photovoltaic conversion efficiency, low cost and stable performance. The deposition methods and preparation conditions for the fabrication of CdTe are very important for the achievement of high solar cell conversion efficiency. There are some rearranged reports about the deposition methods available for the preparation of CdTe thin films such as close spaced sublimation (CSS), physical vapor deposition (PVD), vacuum evaporation, vapor transport deposition (VTD), closed space vapor transport, electrodeposition, screen printing, spray pyrolysis, metalorganic chemical vapor deposition (MOCVD), and RF sputtering. The RF sputtering method for the preparation of CdTe thin films has important advantages in that the thin films can be prepared at low growth temperatures with large-area deposition suitable for mass-production. The authors reported that the optical and electrical properties of CdTe thin film were closely connected by the thickness-uniformity of the film in the previous study [1], which means that the better optical absorbance and the higher carrier concentration could be obtained in the better condition of thickness-uniformity for CdTe thin film. The thickness-uniformity could be controlled and improved by the some process parameters such as vacuum level and RF power in the sputtering process of CdTe thin films. However, there is a limitation to improve the thickness-uniformity only in the preparation process [1]. So it is necessary to introduce the external or additional method for improving the thickness-uniformity of CdTe thin film because the cell size of thin film solar cell will be enlarged. Therefore, the authors firstly applied the chemical mechanical polishing (CMP) process to improving the thickness-uniformity of CdTe thin films with a G&P POLI-450 CMP polisher [2]. CMP process is the most important process in semiconductor manufacturing processes in order to planarize the surface of the wafer even over 300 mm and to form the copper interconnects with damascene process. Some important CMP characteristics for CdTe were obtained including removal rate (RR), WIWNU%, RMS roughness, and peak-to-valley roughness [2]. With these important results, the CMP process for CdTe thin films was performed to improve the thickness-uniformity of the sputtering-deposited CdTe thin film which had the worst two thickness-uniformities of them. Some optical properties including optical transmittance and absorbance of the CdTe thin films were measured by using a UV-Visible spectrophotometer (Varian Techtron, Cary500scan) in the range of 400 - 800 nm. After CMP process, the thickness-uniformities became better than that of the best condition in the previous sputtering process of CdTe thin films. Consequently, the optical properties were directly affected by the thickness-uniformity of CdTe thin film. The absorbance of CdTe thin films was improved although the thickness of CdTe thin film was not changed.

• Korean Journal of Poultry Science
• /
• v.44 no.1
• /
• pp.29-39
• /
• 2017

This study examined the effects of adding dried blood mixed with Macsumsuk to the feed of broiler chickens. The blood had been dried at $200^{\circ}C$ in an ultra-high-temperature injection system and mixed in a 70:30 proportion of blood meal to Macsumsuk. The experiment consisted of four treatment groups of 150 chickens each. The control group received common broiler feed only, while treatment groups T1, T2, and T3 received feed supplemented with 0.5, 1.0, and 3.0% of the blood meal/Macsumsuk mixture, respectively. The diets were fed for a total of 35 days. Compared with the controls, body weight gain was improved in groups T1 (1,621 g), T2 (1,749 g), and T3 (1,739 g) (1,621-1,749 g vs. 1,448.5 g, respectively) and feed efficiency increased (p<0.01). The carcass rate in group T3 was higher by 83.26% than that in the controls (75.96 %) (p<0.01). The water holding capacity (WHC) increased in groups T1 and T2 (62.27 and 63.80% respectively) compared with controls (p<0.01). The intestine length was longer in groups T1 and T2 (53.98-55.48) than in controls (45.81) (p<0.01). Adding 0.5-1.0% of the dried blood meal Macsumsuk supplement resulted in a significant reduction in the cholesterol content (39.28~47.34 mg/100 g) compared with the controls (50.44 mg/100 g) (p<0.001); furthermore, the proportions of fatty acids including oleic (C18:1n9), ${\gamma}-linoleic$ (C18:3n6), eicosenoic (C20:1n9), and arachidonic (C20:4n6) acids were significantly increased (p<0.01). Compared with controls, the proportion of unsaturated fatty acids (UFA), mono-unsaturated fatty acids (MUFA), and proportion of UFA / SFA in group T1 was (68.66 vs. 69.35%, 51.22 vs 52.00%, and 2.19 vs. 2.26%, respectively) (p<0.05). However, the amino acid content of cystine and methionine of the treatment group (0.43~0.57%) was significantly higher than that of controls (0.38~0.46%) (p<0.05). Overall, supplementing the feed with 0.5-1.0% of the mixture of blood meal Macsumsuk improved productivity by increasing weight gain and feed efficiency, improved meat quality by increasing the water-holding capacity and levels of unsaturated fatty acids, and improved meat color.

• Korean Journal of Organic Agriculture
• /
• v.15 no.2
• /
• pp.195-205
• /
• 2007

This study was conducted to determine effects of probiotics on performance, nutrient digestibility and nitrogen retention in Korean black goats fed two diets differing in forage to concentrate (F:C) ratio (30:70 and 70:30), for the establishment of their more efficient feeding management system. The probiotics employed in this trial were mixtures of different microbial species, which consisted of Lactobacillus casei, Bacillus subtilis, Saccharomyces cerevisiae, Aspergillus oryzae and Streptomyces griseus. Additional levels of probiotics to each F:C ratio (70:30 or 30:70) were 0 and 0.2%, respectively. Thus, twelve Korean black male goats were allotted to treatments in four groups of three goats per treatment and then they were housed in individual metabolism cages with a randomized complete block design for 21 days. Dry matter (DM) intakes were not affected by dietary F:C ratio and probiotics. Digestible DM amounts were significantly (p<0.05) decreased with increasing levels of dietary forage, but they were not affected by probiotics addition. Dry matter intakes per metabolic body weight and their ratio per body weight had a similar trend to DM intakes with no significant difference by F:C ratio and probiotics addition. The nutrient digestibility was significantly (p<0.05) increased with decreasing levels of forage in the diet, but it was not affected by probiotics addition. Urinary nitrogen loss was significantly (p<0.05) decreased with decreasing levels of dietary forage, but there was no significant difference between probiotics-supplemented and unsupplemented groups within the same F:C ratio. On the contrary, nitrogen retention was decreased with increasing levels of dietary forage, and probiotics supplementation to two diets differing in F:C ratio showed slightly increasing trends in the nitrogen retention. From the above results, probiotics supplementation to two diets differing in F:C ratio did not have the significant influence on feed intakes, nutrient digestibility and nitrogen retention. Consequently, these parameters of Korean black goats were dominated rather by F:C ratio than by dietary probiotics.

• Food Science of Animal Resources
• /
• v.23 no.1
• /
• pp.9-15
• /
• 2003

All diets were based on feeds of fattening period pigs(LY x D, ca. 90 kg) with six treatments, which were the control, containing 5% beef tallow(C), 3% beef tallows and 2% perilla seeds oil(T1), 250 ppm vit. E(${\alpha}$-tocopheryl acetate) in T1(T2), 3% beef tallow and 2% squid viscera oil(T3), 250 ppm vit. E in T3(T4), and 3% beef tallow and 2% CLA(T5), respectively. Produced porks and their carcass characteristics were as follows. The daily gain of pigs was higher in T2 and 73 than any other treatments(p<0.05). Its T2 and T3 was 3.71 and 3.80 respectively, however, there was no significance in feed intake. The highest back fat thickness was shown in control group on market weight, while there was no significant difference on their initial weight. Loin-eye muscle area did not show any significant difference on initial weight and on market weight, however, its T5 was about twice as large as T2's. Content of triglyceride in blood was high in control group as compared to others; especially, the values for T3, T4 and T5 were significantly low(p <0.05). There was no significant difference in total cholesterol contents, and the ratio of HDL cholesterol/total cholesterol was higher in vit. E treated samples than untreated sample. Atherogenic index was high in sample with T3 and low in sample with T2. The perilla seed oil, squid fish oil, and vit. E decreased atherogenic index. Dressing percentage, back fat thickness, and grade did not show any significant difference(p >0.05); however, T2, C and T3, T1 and T5 showed 4.67, 4.29, 4.00 respectively, in grades.

• Asian Pacific Journal of Cancer Prevention
• /
• v.15 no.4
• /
• pp.1811-1815
• /
• 2014

Background: In clinical trials with no upper age limit, the proportion of older patients is usually small, probably reflecting the more conservative approach adopted by clinicians when treating the elderly. An exploratory analysis of elderly patients in the RECORD-1 Trial showed that patients ${\geq}$ 65 y.o. had superior median PFS than overall RECORD-1 population (5.4 months and 4.9 months, respectively). We investigated the efficacy, relative benefit and safety of Everolimus (EVE) as sequential therapy after failure of VEGFr-TKI therapy for older patients with metastatic renal cell cancer (mRCC), in daily practice. Materials and Methods: 172 consecutive IRB approved patients with mRCC (median age 65, M:F 135/37, 78% clear cell) who received salvage EVE at 39 tertiary institutions between October 2009 and August 2011 were included in this analysis. Some 31% had progressed on sunitinib, 22% on sorafenib, 1% on axitinib, 41% on sequential therapy, and 5% had received other therapy. Patients with brain metastases were not included and 95% of the patients had a ECOG (Eastern Cooperative Oncology Group) performance status (PS) of 0 or 1. Previous radiotherapy was an exclusion criterion, but prior chemotherapy was permitted. Adequate organ function and hematologic parameters were mandatory. EVE administration was approved by the institutional review board at each participating institution and signed informed consent was obtained from all patients. Results: Median time of the whole cohort to last follow-up was 3.5 months (range 0.4-15.2 months). Forty four percent were continuing to take EVE at last followup. There were 86 (50%) patients ${\geq}$ 65 y.o. and 86 (50%) <65 y.o. The percentage of patients who showed PR/SD was higher in the older group than in the younger one (5.9%/61.2% vs 1.2%/46.5%, respectively). Median survival of older patients was also significantly longer (3.5 +/- 0.31 vs 3.1 +/- 0.34, hazard ratio=0.45, CI; 0.255-0.802). Analysis using Cox regression model adjusted for gender, PS, number of metastases, site of metastases, histology, smoking history and age detected an association between age and PFS (p=0.011). The frequency of adverse events in elderly patients treated with EVE was no greater than that in younger patients, although such toxicity may have had a greater impact on their quality of life. Conclusions: Older patients should not generally be excluded from accepted therapies (mTOR inhibitors after failure of VEGFr-TKI therapy) for mRCC.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.54 no.5
• /
• pp.133-137
• /
• 2017

The purpose of this study was to validate for GATE (Geant4 Application for Tomographic Emission) simulation by comparing the results of GATE simulation and experiment in real SPECT system. Futhermore, we want to prove that it is possible that the quantitative research of gamma camera/SPECT imaging for therapeutic radio isotope by using GATE simulation. In this study, the SPECT system on simulation referred to the parameters of Stream-R Forte version 1.2 (Philips Medical System, Best and Heerlen, Netherlands). To understand the I-131 image of gamma camera/SPECT system, we acquired the energy spectrum and measured the full width at half maximum (FWHM) which comes from line spread function (LSF) with and without scatter material in real SPECT system. And to compare with experiment, we also measured the FWHM and acquired the energy spectrum without scatter material in GATE simulation. As a result, without scatter material, the energy peak was almost same location, which are located nearby 364 keV, and other spectrum factors are same tendency in both cases. The FWHM was increased by increasing the distance of source to detector, and the error rate was approximately 3.8%. When we used the line source with scatter material, energy spectrum also indicated similar tendency in both cases. As you confirmed earlier, GATE simulation included real instrument and radioisotope characters for therapeutic radioisotope. Therefore this result that it was possible that various quantitative study for therapeutic radioisotope imaging in gamma camera/SPECT using GATE simulation.