• MALSORI
• /
• no.65
• /
• pp.93-103
• /
• 2008

In this paper, we propose a method for optimizing a multiple pronunciation dictionary used for modeling pronunciation variations of non-native speech. The proposed method removes some confusable pronunciation variants in the dictionary, resulting in a reduced dictionary size and less decoding time for automatic speech recognition (ASR). To this end, a confusability measure is first defined based on the Levenshtein distance between two different pronunciation variants. Then, the number of phonemes for each pronunciation variant is incorporated into the confusability measure to compensate for ASR errors due to words of a shorter length. We investigate the effect of the proposed method on ASR performance, where Korean is selected as the target language and Korean utterances spoken by Chinese native speakers are considered as non-native speech. It is shown from the experiments that an ASR system using the multiple pronunciation dictionary optimized by the proposed method can provide a relative average word error rate reduction of 6.25%, with 11.67% less ASR decoding time, as compared with that using a multiple pronunciation dictionary without the optimization.

• Asia Pacific Journal of Corpus Research
• /
• v.4 no.1
• /
• pp.1-23
• /
• 2023

The Japanese language, as a typical agglutinating language, permits large noun phrases (NP) containing ten or more morphemes. In this paper, we argue that the nature of the NP in Japanese is changing. Our data are drawn from the Balanced Corpus of Contemporary Written Japanese. We conduct a series of apparent-time studies of ongoing changes in complex NPs. We first examine the length of compound nouns, followed by the usage of bound suffixes. We then examine ongoing changes in complex NPs that contain genitive case markers. Finally, we examine noun incorporation. All of our studies show a trend towards shorter, less complex NPs. Furthermore, our results suggest that the usage rate of phrases that modify the noun inside the NP (compound nouns, bound nouns, NPs containing genitive case, noun incorporation) appears to be decreasing over time. On the other hand, the usage rate of modifying material outside of the NP (positional phrases, relative clauses) appears to be increasing over time. We conclude by suggesting that our results reflect a diachronic change of decreasing synthetic morphology and increasing analytic morphology. We end by pointing out the implications of this work on our understanding syntheticity and analyticity.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.17 no.2
• /
• pp.397-404
• /
• 2013

This paper propose the hardware architecture of face detection hardware system using the AdaBoost algorithm. The proposed structure of face detection hardware system is possible to work in 30frame per second and in real time. And the AdaBoost algorithm is adopted to learn and generate the characteristics of the face data by Matlab, and finally detected the face using this data. This paper describes the face detection hardware structure composed of image scaler, integral image extraction, face comparing, memory interface, data grouper and detected result display. The proposed circuit is so designed to process one point in one cycle that the prosed design can process full HD($1920{\times}1080$) image at 70MHz, which is approximate $2316087{\times}30$ cycle. Furthermore, This paper use the reducing the word length by Overflow to reduce memory size. and the proposed structure for face detection has been designed using Verilog HDL and modified in Mentor Graphics Modelsim. The proposed structure has been work on 45MHz operating frequency and use 74,757 LUT in FPGA Xilinx Virtex-5 XC5LX330.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.39 no.3
• /
• pp.90-99
• /
• 2016

Forecasting of box office performance after a film release is very important, from the viewpoint of increase profitability by reducing the production cost and the marketing cost. Analysis of psychological factors such as word-of-mouth and expert assessment is essential, but hard to perform due to the difficulties of data collection. Information technology such as web crawling and text mining can help to overcome this situation. For effective text mining, categorization of objects is required. In this perspective, the objective of this study is to provide a framework for classifying films according to their characteristics. Data including psychological factors are collected from Web sites using the web crawling. A clustering analysis is conducted to classify films and a series of one-way ANOVA analysis are conducted to statistically verify the differences of characteristics among groups. The result of the cluster analysis based on the review and revenues shows that the films can be categorized into four distinct groups and the differences of characteristics are statistically significant. The first group is high sales of the box office and the number of clicks on reviews is higher than other groups. The characteristic of the second group is similar with the 1st group, while the length of review is longer and the box office sales are not good. The third group's audiences prefer to documentaries and animations and the number of comments and interests are significantly lower than other groups. The last group prefer to criminal, thriller and suspense genre. Correspondence analysis is also conducted to match the groups and intrinsic characteristics of films such as genre, movie rating and nation.

• Journal of the Korean Society of Costume
• /
• v.67 no.3
• /
• pp.15-30
• /
• 2017

The purpose of this study is to analyze the formative characteristics of Hanbok among youngsters based on SNS proof shots, identify new characteristics of Hanbok as part of play and travel rather than as formal Hanbok, and provide information for the Hanbok market. As research methodology, our search was carried out by using '#Hanbok Travel' as the search word in Instagram, where the Hanbok proof shot phenomenon is actively under way. A total of 535 posts from March 21, 2016 to April 1, 2016 were selected as objects of this study, excluding posts containing Hanbok with indiscernible shape, Korean traditional costume manufacturers' promotional posts, and repetitive posts by one person. First, the 535 posts were analyzed by season, region, number of people, and gender, and after men's data were excluded, 644 Hanboks were left for analysis. Their formative characteristics were analyzed by using SPSS 21.0. The results showed that the formative characteristics of Hanbok shown in SNS proof shots included diversification of length in jeogori(Korean traditional jacket), skirt, and sleeve, use of pragmatic material and achromatic color, and reduced use of decorative technique. Hanboks shown in the Hanbok proof shots should be considered as significant data because each shots show clothes selected and worn directly by user's side, unlike the existing studies centering on Hanbok designers' works.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.12 no.4
• /
• pp.309-324
• /
• 1987

In recent years has been an increasing interest with respect to using the discrete cosine transform(DCT) of which performance is found close to that of the Karhumen-Loeve transform, known to be optimal in the area of digital image processing for tha purpose of the image data compression. Among most of reported algorithms aimed at lowering the coputation complexity. Chen's algorithm is is found to be most popular, Recently, Lee proposed a now algorithm of which the computational complexity is lower than that of Chen's. but its performance is significantly degraded by FWL(Finite-Word-Lenght) effects as a result of employinga a fixed-poing arithmetic. In this paper performance evaluation of these two algorithms and error analysis of FWL effect are described. Also a scaling technique which we call Up & Down-scaling is proposed to allevaiate a performance degradation due to fixed-point arithmetic. When the 16x16point 2DCT is applied on image data and a 16-bit fixed-point arithmetic is employed, both the analysis and simulation show that is colse to that of Chen's.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.19 no.6
• /
• pp.773-778
• /
• 2009

This paper studied the training methods less affected by the emotional variation for the development of the robust speech recognition system. For this purpose, the effect of emotional variations on the speech signal were studied using speech database containing various emotions. The performance of the speech recognition system trained by using the speech signal containing no emotion is deteriorated if the test speech signal contains the emotions because of the emotional difference between the test and training data. In this study, it is observed that vocal tract length of the speaker is affected by the emotional variation and this effect is one of the reasons that makes the performance of the speech recognition system worse. In this paper, vocal tract normalization method is used to develop the robust speech recognition system for emotional variations. Experimental results from the isolated word recognition using HMM showed that the vocal tract normalization method reduced the error rate of the conventional recognition system by 41.9% when emotional test data was used.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.20 no.4
• /
• pp.528-533
• /
• 2010

This paper studied the training methods less affected by the emotional variation for the development of the robust speech recognition system. For this purpose, the effect of emotional variation on the speech signal and the speech recognition system were studied using speech database containing various emotions. The performance of the speech recognition system trained by using the speech signal containing no emotion is deteriorated if the test speech signal contains the emotions because of the emotional difference between the test and training data. In this study, it is observed that vocal tract length of the speaker is affected by the emotional variation and this effect is one of the reasons that makes the performance of the speech recognition system worse. In this paper, a training method that cover the speech variations is proposed to develop the emotionally robust speech recognition system. Experimental results from the isolated word recognition using HMM showed that propose method reduced the error rate of the conventional recognition system by 28.4% when emotional test data was used.

• Journal of the Korean Home Economics Association
• /
• v.8 no.1
• /
• pp.62-86
• /
• 1970

A study on the upper garment of Korean women, JUORI The upper garment of Korean women. JUGORI, is an inherited mode from the ancient clothing style in the various aspects based on the particulars of Korean clothes. The ancient style of clothes is originated from KWAMDUI belonging to inhabitants of Northern Territory of Korea. And it is quite different from Chinese clothes in lineage. However, this unicque mode of clothes has been much influnced by the Chinese culture and also by the climate of Korea. And it is quite different from Chinese clothes in lineage. However, this unicque mode of clothes has been much influnced by the Chinese culture and also by the climate of Korean penynsula. The changes of the pattern of JUGORI, in a word, is a sign of shortening tendency of size. This tendency of JUGORI is remarkably seen in the shortening of length and other parts are decreased in size. The JUGORI in the ancient age was fallen below the weist of woman, which is similar to Robe, and was worn with band. However, the length of the JUGORI has been gradually shortened, and therefore, GORUM took place of the band. The shortening tendency of JUGORI is seemed to be shown its sign in the initial time of its origin, because there are some evidences that the women in Sylla Dynasty, and this tendency has been much expedited during the period of Koryu Dynasty with influences of Monggorian culture (Won Lynasty of China) The oldest sample for data of JUGORI in nowaday is one the remains of Yi Dynasty, and this sample for data provides all the particulars of the modern pattern of JUGORI. The tendency of JUGORI had been continued even in Yi Dynasty, and at the end of the Dynasty, the clothes was shortened that the women felt inconvenient wearing it in the status of the shortened JUGORI which was even hardly cover the initial time of epoch of modernization induced from the Western civilization, and after 1920s and 1930s JUGORI become a larger tendency. This is a sing of revival of practical use and rationalization of JUGORI become a shortening tendency again, and the size is similar with that of early age of Yi Dynasty. Instead of these similarities, the particulars of modern JUGORI is weighing on much emphasis on curve beauty and expression of experior beauty. The reason is that, together with westernization of clothes, JUGORI became a special pattern of clothes as a traditional Korean women wears. The very thing explaining this pattern of JUGORI is the "ARIRANG DRESS". And there are some fashion using button instead of GORUM and half sleeve JUGORI for summer use which is regarded as a part of improved aspect of life in Korea. in Korea.

• Journal of IKEEE
• /
• v.8 no.2 s.15
• /
• pp.186-194
• /
• 2004

Due to the exponential growth of information on the internet, it is getting difficult to find and organize relevant informations. To reduce heavy overload of accesses to information, automatic text classification for handling enormous documents is necessary. In this paper, we describe structure and implementation of a document classification system for web documents. We utilize SVM for documentation classification model that is constructed based on training set and its representative terms in a directory. In our system, SVM is trained and is used for document classification by using word set that is extracted from information and communication related web documents. In addition, we use vector-space model in order to represent characteristics based on TFiDF and training data consists of positive and negative classes that are represented by using characteristic set with weight. Experiments show the results of categorization and the correlation of vector length.