• Title/Summary/Keyword: 합성성과지표

Search Result 13, Processing Time 0.033 seconds

One-shot multi-speaker text-to-speech using RawNet3 speaker representation (RawNet3를 통해 추출한 화자 특성 기반 원샷 다화자 음성합성 시스템)

  • Sohee Han;Jisub Um;Hoirin Kim
    • Phonetics and Speech Sciences
    • /
    • v.16 no.1
    • /
    • pp.67-76
    • /
    • 2024
  • Recent advances in text-to-speech (TTS) technology have significantly improved the quality of synthesized speech, reaching a level where it can closely imitate natural human speech. Especially, TTS models offering various voice characteristics and personalized speech, are widely utilized in fields such as artificial intelligence (AI) tutors, advertising, and video dubbing. Accordingly, in this paper, we propose a one-shot multi-speaker TTS system that can ensure acoustic diversity and synthesize personalized voice by generating speech using unseen target speakers' utterances. The proposed model integrates a speaker encoder into a TTS model consisting of the FastSpeech2 acoustic model and the HiFi-GAN vocoder. The speaker encoder, based on the pre-trained RawNet3, extracts speaker-specific voice features. Furthermore, the proposed approach not only includes an English one-shot multi-speaker TTS but also introduces a Korean one-shot multi-speaker TTS. We evaluate naturalness and speaker similarity of the generated speech using objective and subjective metrics. In the subjective evaluation, the proposed Korean one-shot multi-speaker TTS obtained naturalness mean opinion score (NMOS) of 3.36 and similarity MOS (SMOS) of 3.16. The objective evaluation of the proposed English and Korean one-shot multi-speaker TTS showed a prediction MOS (P-MOS) of 2.54 and 3.74, respectively. These results indicate that the performance of our proposed model is improved over the baseline models in terms of both naturalness and speaker similarity.

The Economic Impact of the May 18 Democratic Uprising on the Regional Economy: A Synthetic Control Method (SCM) approach (5·18민주화운동이 지역경제에 미친 경제적 영향 분석: 통제집단합성법(SCM)을 이용한 접근)

  • Ryu, Deockhyun;Seo, Dongkyu
    • Analyses & Alternatives
    • /
    • v.6 no.2
    • /
    • pp.155-183
    • /
    • 2022
  • The purpose of this study is to econometrically analyze the negative impact of the May 18 Democratic Uprising on the Gwangju/Jeonnam regionional economy using the Synthetic Control Method (SCM). The SCM SCM is a methodology similar to the difference-in-difference(DID) method of microeconometrics. It is applied to macroeconomic variables such as country, region, etc. to estimate the causal relationship between specific events and the dependent variable. In this study, as of 1980, local tax revenue data of metropolitan local governments were used as a proxy variable for the economy of the region, and the impact of the May 18 Democratic Uprising on the economy of Gwangju/Jeonnam region was analyzed through various socio-economic indicators. In this study, data were used to analyze from 1971 to 2000, and as a result of empirical analysis, local tax revenues in Gwangju/Jeonnam area were less collected than normal routes up to 17%. In addition, the significance of this analysis was confirmed through in-time placebo effect analysis and in-space placebo effect analysis, which are methods of analyzing the robustness of the control group synthesis method.

Uncertainty Assessment of Radar Reflectivity-Rainfall Relationship based on Bayesian Perspective using Long-term Radar Reflectivity (장기간 레이더 반사도를 활용한 Bayesian 추론 기반의 레이더 반사도-강수량 관계식 불확실성 평가)

  • Kim, Tae-Jeong;Kim, Ho Jun;Kwon, Hyun-Han
    • Proceedings of the Korea Water Resources Association Conference
    • /
    • 2020.06a
    • /
    • pp.61-61
    • /
    • 2020
  • 최근 수문기상학 분야에서 레이더 강수량을 활용한 응용연구가 활발하게 진행되고 있다. 하지만 레이더 강수량은 대류성 및 층상형 등과 같이 강수특성을 기준으로 레이더 반사도-강수량(Reflectivity-Rainfall, Z-R) 관계식 매개변수를 시공간적으로 동일하게 적용하여 레이더 강수량을 산정하는 방법론은 지상관측 강수량과 정량적인 편의 오차(systematic error)를 발생시킬 수 있는 문제점이 있다. 본 연구는 장기간의 레이더 합성장 반사도를 활용하여 Z-R 관계식 매개변수를 산정하였으며, 이 과정에서 Bayesian 추론 기법을 도입하여 Z-R 관계식 매개변수의 불확실성을 정량화하였다. 추가적으로 편의 오차를 최소화하기 위하여 계절성을 고려한 Z-R 관계식을 산정하였다. 건기와 우기로 구분하여 산정된 Z-R 관계식 매개변수의 공간적으로 변동성과 더불어 강수의 계절적 특성에 기인하는 Z-R 관계식 매개변수의 역비례 관계를 확인하였다. 최종적으로, 제안된 방법론으로 산정된 레이더 강수장은 일반적으로 레이더 강수량 산정에 널리 이용되는 Marshall-Palmer Z-R 관계식으로 산정된 강수장에 비하여 우수한 통계지표를 제시하였다.

  • PDF

A Lower Bound Estimation on the Number of Micro-Registers in Time-Multiplexed FPGA Synthesis (시분할 FPGA 합성에서 마이크로 레지스터 개수에 대한 하한 추정 기법)

  • 엄성용
    • Journal of KIISE:Computer Systems and Theory
    • /
    • v.30 no.9
    • /
    • pp.512-522
    • /
    • 2003
  • For a time-multiplexed FPGA, a circuit is partitioned into several subcircuits, so that they temporally share the same physical FPGA device by hardware reconfiguration. In these architectures, all the hardware reconfiguration information called contexts are generated and downloaded into the chip, and then the pre-scheduled context switches occur properly and timely. Typically, the size of the chip required to implement the circuit depends on both the maximum number of the LUT blocks required to implement the function of each subcircuit and the maximum number of micro-registers to store results over context switches in the same time. Therefore, many partitioning or synthesis methods try to minimize these two factors. In this paper, we present a new estimation technique to find the lower bound on the number of micro-registers which can be obtained by any synthesis methods, respectively, without performing any actual synthesis and/or design space exploration. The lower bound estimation is very important in sense that it greatly helps to evaluate the results of the previous work and even the future work. If the estimated lower bound exactly matches the actual number in the actual design result, we can say that the result is guaranteed to be optimal. In contrast, if they do not match, the following two cases are expected: we might estimate a better (more exact) lower bound or we find a new synthesis result better than those of the previous work. Our experimental results show that there are some differences between the numbers of micro-registers and our estimated lower bounds. One reason for these differences seems that our estimation tries to estimate the result with the minimum micro-registers among all the possible candidates, regardless of usage of other resources such as LUTs, while the previous work takes into account both LUTs and micro-registers. In addition, it implies that our method may have some limitation on exact estimation due to the complexity of the problem itself in sense that it is much more complicated than LUT estimation and thus needs more improvement, and/or there may exist some other synthesis results better than those of the previous work.

Detectability Evaluation for Alert Sound in an Electric Vehicle (전기자동차의 경고음에 대한 인지성 평가)

  • Han, Man Uk;Lee, Sang Kwon
    • Transactions of the Korean Society of Mechanical Engineers A
    • /
    • v.41 no.10
    • /
    • pp.923-929
    • /
    • 2017
  • Generally, the sound emitted from a vehicle powered by an electric motor is lower than that of internal combustion engine vehicles. Therefore, pedestrians often cannot detect approaching electric vehicles. Therefore, a certain additional warning sound is required for these types of automobiles. In this study, to develop an audible warning sound, nine warning sounds are designed based on signal processing and chord theory. The background noise measured on the road is also added to these synthetic sounds. The detectability of these warning sounds is evaluated by subjective tests. The sound metric is correlated to detectability and is investigated through psychoacoustic theory and subjective evaluation. It is determined that known psychoacoustic parameters such as loudness, sharpness, and roughness have a low correlation with detectability. However, it is found that the interval of harmonic sound correlates well with detectability.

Compatibility of MODIS Vegetation Indices and Their Sensitivity to Sensor Geometry (MODIS 식생지수에 미치는 센서 geometry의 영향과 센서 간 자료 호환성 검토)

  • Park, Sunyurp
    • Journal of the Korean Geographical Society
    • /
    • v.49 no.1
    • /
    • pp.45-56
    • /
    • 2014
  • Data composite methods have been typically applied to satellite-based vegetation index(VI) data to continuously acquire vegetation greenness over the land surface. Data composites are useful for construction of long-term archives of vegetation indices by minimizing missing data or contamination from noise. In addition, if multi-sensor vegetation indices that are acquired during the same composite periods are used interchangeably, data stability and continuity may be significantly enhanced. This study evaluated the influences of sensor geometry on MODIS vegetation indices and investigated data compatibility of two difference vegetation indices, the Normalized Difference Vegetation Index(NDVI) and the Enhanced Vegetation Index(EVI), for potential improvement of long-term data construction. Relationships between NDVI and EVI turned out statistically significant with variations among vegetation covers. Due to their curvilinear relationships, NDVI became saturated and leveled off as EVI reached high ranges. Correlation coefficients between Terra- and Aqua-based vegetation indices ranged from 0.747 to 0.963 for EVI, and from 0.641 to 0.880 for NDVI, showing better compatibility for EVI compared to NDVI. In-depth analyses of VI outliers that deviated from regression equations constructed from the two different sensors remain as a future study to improve their compatibility.

  • PDF

A Study on the Modification of NH4+Y-zeolite for Improving Adsorption/Desorption Performance of Benzene (NH4+Y-zeolite의 개질을 통한 벤젠 흡·탈착 성능 증진 연구)

  • Jang, Young Hee;Noh, Young Il;Lee, Sang Moon;Kim, Sung Su
    • Clean Technology
    • /
    • v.25 no.1
    • /
    • pp.33-39
    • /
    • 2019
  • A.C (activated carbon) is mainly used to remove VOCs (volatile organic compounds), however, it has many problems such as fire risk due to increasing of adsorbent surface temperature during VOCs ad/desorption, increased cost by frequent replacement cycles requirement and performance degradation when containing moisture. In order to solve these problems, many researches, hydrophobic zeolite adsorbents, have been reported. In this study, $NH_4{^+}Y$-zeolite was synthesized with Y-zeolite through steam treatment and acid treatment, which is one of the hydrophobic modification methods, to secure high surface area, thermal stability and humidity resistance. The Y, Y-550-HN, Y-600-HN and Y-650-HN had adsorption capacities of $23mg\;g^{-1}$, $38mg\;g^{-1}$, $77mg\;g^{-1}$, $61mg\;g^{-1}$. The change of Si/Al ratio, which is an index to confirm the degree of modification, was confirmed by XRF (X-ray fluorescence spectrometer) analysis. As a result, the adsorbtion performance was improved when Y-zeolite modified, and the Si/Al ratio of Y, Y-550-HN, Y-600-HN, Y-650-HN were increased to 3.1765, 6.6706, 7.3079, and 7.4635, respectively. Whereas it was confirmed that structural crystallization due to high heat treatment temperature affected performance degradation. Therefore, there is an optimal heat treatment temperature of Y-zeolite, optimum modification condition study could be a substitute for activated carbon as a condition for producing an adsorbent having high durability and stability.

A Time Series Graph based Convolutional Neural Network Model for Effective Input Variable Pattern Learning : Application to the Prediction of Stock Market (효과적인 입력변수 패턴 학습을 위한 시계열 그래프 기반 합성곱 신경망 모형: 주식시장 예측에의 응용)

  • Lee, Mo-Se;Ahn, Hyunchul
    • Journal of Intelligence and Information Systems
    • /
    • v.24 no.1
    • /
    • pp.167-181
    • /
    • 2018
  • Over the past decade, deep learning has been in spotlight among various machine learning algorithms. In particular, CNN(Convolutional Neural Network), which is known as the effective solution for recognizing and classifying images or voices, has been popularly applied to classification and prediction problems. In this study, we investigate the way to apply CNN in business problem solving. Specifically, this study propose to apply CNN to stock market prediction, one of the most challenging tasks in the machine learning research. As mentioned, CNN has strength in interpreting images. Thus, the model proposed in this study adopts CNN as the binary classifier that predicts stock market direction (upward or downward) by using time series graphs as its inputs. That is, our proposal is to build a machine learning algorithm that mimics an experts called 'technical analysts' who examine the graph of past price movement, and predict future financial price movements. Our proposed model named 'CNN-FG(Convolutional Neural Network using Fluctuation Graph)' consists of five steps. In the first step, it divides the dataset into the intervals of 5 days. And then, it creates time series graphs for the divided dataset in step 2. The size of the image in which the graph is drawn is $40(pixels){\times}40(pixels)$, and the graph of each independent variable was drawn using different colors. In step 3, the model converts the images into the matrices. Each image is converted into the combination of three matrices in order to express the value of the color using R(red), G(green), and B(blue) scale. In the next step, it splits the dataset of the graph images into training and validation datasets. We used 80% of the total dataset as the training dataset, and the remaining 20% as the validation dataset. And then, CNN classifiers are trained using the images of training dataset in the final step. Regarding the parameters of CNN-FG, we adopted two convolution filters ($5{\times}5{\times}6$ and $5{\times}5{\times}9$) in the convolution layer. In the pooling layer, $2{\times}2$ max pooling filter was used. The numbers of the nodes in two hidden layers were set to, respectively, 900 and 32, and the number of the nodes in the output layer was set to 2(one is for the prediction of upward trend, and the other one is for downward trend). Activation functions for the convolution layer and the hidden layer were set to ReLU(Rectified Linear Unit), and one for the output layer set to Softmax function. To validate our model - CNN-FG, we applied it to the prediction of KOSPI200 for 2,026 days in eight years (from 2009 to 2016). To match the proportions of the two groups in the independent variable (i.e. tomorrow's stock market movement), we selected 1,950 samples by applying random sampling. Finally, we built the training dataset using 80% of the total dataset (1,560 samples), and the validation dataset using 20% (390 samples). The dependent variables of the experimental dataset included twelve technical indicators popularly been used in the previous studies. They include Stochastic %K, Stochastic %D, Momentum, ROC(rate of change), LW %R(Larry William's %R), A/D oscillator(accumulation/distribution oscillator), OSCP(price oscillator), CCI(commodity channel index), and so on. To confirm the superiority of CNN-FG, we compared its prediction accuracy with the ones of other classification models. Experimental results showed that CNN-FG outperforms LOGIT(logistic regression), ANN(artificial neural network), and SVM(support vector machine) with the statistical significance. These empirical results imply that converting time series business data into graphs and building CNN-based classification models using these graphs can be effective from the perspective of prediction accuracy. Thus, this paper sheds a light on how to apply deep learning techniques to the domain of business problem solving.

The Amount of Telomeric DNA and Telomerase Activity on Cattle Cells (소의 생리적 특성에따름 세포내 텔로미어 함량과 텔로머레이스 활성도 분석)

  • Choi, Duk-Soon;Cho, Chang-Yeon;Sohn, Sea-Hwan
    • Journal of Animal Science and Technology
    • /
    • v.50 no.4
    • /
    • pp.445-456
    • /
    • 2008
  • Telomeres consist of TTAGGG tandem repeated DNA sequences with specific proteins and locate at chromosome ends. Telomeres are essential for chromosome stability and are related with cell senescence, apoptosis and cancer. Telomerase is a ribonucleoprotein which has a template for the synthesis of telomeric DNA. This study was carried out to analyze the amount of telomeric DNA and telomerase activity in cattle cells. Analysis of the quantity of telomere in lymphocytes was done at different ages, sex and among Korean cattle and Holstein breeds. The telomerase activity was also analyzed in liver, brain, heart, kidney, and testis tissues of fetal calf and of 18 month old cattle. The amount of telomeres in lymphocytes and other tissue cells was analyzed by Quantitative-Fluorescence in situ Hybridization (Q-FISH) technique using a telomeric DNA probe. Telomerase activity was analyzed by Telomeric Repeat Amplification Protocol assay (TRAP). The amount of telomeric DNA on the lymphocytes during the whole life span was decreased along with age. Quantity of telomeres in Korean cattle was significantly higher than that in Holstein breed. The amount of telomeric DNA in males was significantly higher than that in females. Telomerase activity was up-regulated in most bovine tissues during fetal stage, but was down-regulated in most tissues at mature 18 month age except the testis cells. This study indicates that the amount of telomeres and telomerase activity of cells can be used as an age marker or/and a physiological marker of cattle.

Study on the Compared between u-Learning and e-Learning based SCORM (SCORM 기반 u-Learning과 e-Learning 비교연구)

  • Choi, Sung;Ryu, Gab-Sang
    • Proceedings of the Korea Society of Information Technology Applications Conference
    • /
    • 2006.06a
    • /
    • pp.495-505
    • /
    • 2006
  • IT기술기반 교육시스템은 인터넷 등장 이전에도 가능성을 인정받아 지속적으로 개발되어 온 분야이며, 교육공학과의 연계로 지식 전날의 이론체계로 각광을 받고 있다. 사이버교유도 인터넷이전부터 다양한 통신방법을 응용하여 개발되었고, 최근 인터넷을 통하여 사이버 교육시스템은 완벽한 기술기반을 갖추게 되였다. 그러나 IT기술의 급격한 변화로 사이버교육시스템은 계속하여 신기술 변화에 적용해야만 한다. 현재 정보통신기술의 변화는 방송 통신망의 융합, 브로드 밴드 네트워킹, 스마트 디바이스의 다양화, 멀티미디어 기술의 고도화로 요약된다. 이 기술의 종합한 작용으로 유비쿼터스 사회의 기반으로 진화되고 있다. 그래서 e-Learning 분야도 기존 인터넷기반 시스템과는 달리 차세대 온라인교육시스템으로 친화되고 있다. IT융합가술 기반의 온라인 교육시스템은 각종 국제표준단체에서 표준안이 제시되고 있다. e-Learning 시스템이란 선기술 기반을 반영한 표준기술을 사용하는 온라인교육시스템을 포괄하는 개념이다. 본 연구에서는 e-Learning 시스템과 유비쿼터스 기술을 반영한 e-Learning을 비교하였다. 그리고 u-Learning 시스템의 기술정립과 EOD(Education On Demand) 시스템에 대하여 연구하였다. 1. u-Learning 정의 정보산업분야를 비롯한 문화, 교육 등 모든 분야에서 유비퀴터스라는 수식어가 붙어 다니고 있다. e- Learning 교육 업계에 따르면 10년 후에는 유비쿼터스는 대중화가 될 것이며, 부가가치 규모는 100조 원에 이를 것으로 추정된다. 그래서 교육산업도 주변 환경이 아날로그 방식에서 IT 기반에 의한 디지털 환경으로 변화되고 있다. 또한 e러닝, T러닝, m러닝, u러닝 등의 용어가 생성되고 있다.키지에어컨에서 사용되고 있는 밀폐형 압축기에 대해서 그림 2에서 나타내고 있는 냉방능력 10tons(120,000Btu/h) 이하를 중심으로 상기의 최근 기술 동향을 간략하게 소개하고자 한다.질표준의 지표성분으로 간주되는 진세노사이드의 절대함량과 그 성분조성 차이에 따른 임상효과의 차별성이 있는지에 대한 검토와, 특히 최근 실험적으로 밝혀지고 있는 사포닌 성분의 장내 세균에 의한 생물전환체의 인체 실험을 통한 효과 검정이 필요하다. 나아가서는 적정 복용량의 설정과 이와 관련되는 생체내 동태 및 생체이용율(bioavilability)에 관한 정보가 거의 없으므로 이것도 금후 검토해야 할 과제로 사료된다. 인삼은 전통약물로서 오랜 역사성과 그동안의 연구결과에 의한 과학성을 가지고 있으므로 건강유지와 병의 예방 및 회복촉진을 위한 보조요법제 또는 기능성 식품으로써의 유용성이 있는 것으로 판단된다. 앞으로 인삼의 활용성 증대를 위해서는 보다 과학적인 임상평가에 의한 안전성 및 유효성 입증과 제품의 엄격한 품질관리의 필요성이 더욱 강조되어야 할 것이다.xyl radical 생성 억제 효과를 보여 주었다. 본 실험을 통하여 BHT 를 제외하고 전반적으로 세포 수준에서의 oxidative stress 에 대한 억제 효과를 확인해 볼 수 있었으며 특히 수용성 항산화제들에서 두드러진 효과를 보여 주었다. 제공하여 내수기반 확충에도 노력해야 할 것 이다.있었다., 인삼이 성장될 때 부분적인 영양상태의 불충분이나 기후 등에 따른 영향을 받을 수 있기 때문에 앞으로 이에 대한 많은 연구가 이루어져야할 것으로 판단된다.태에도 불구하고 [-wh]의미의 겹의문사는 병렬적 관계의 합성어가 아니라 내부구조를 지니지 않은 단순한 단어(minimal $X^{0}$<

  • PDF