• Speech Sciences
• /
• v.8 no.4
• /
• pp.99-107
• /
• 2001

The purpose of this study was to investigate the syllable and phoneme frequencies in the spontaneous speech of some Korean children. Sixty four normally developing children aged from 2 to 5 were involved (male: female=1 : 1, 16 children in each age group). Fifty connected utterances were analyzed using the KCLA (Korean Computerized Language Analysis) 2.0 and Exel. The findings were as follows: 1) /i/ was the most frequently used syllable and was followed by /yo/, /k/, /s'/, /nen/ and so on. 2) The most frequently used Korean phonemes were syllable-initial consonant /k/, syllable- medial vowel /a/ and syllable-final consonant /n/. 3) All seven syllable final consonants (/p,t,k,m,n,n,l/) were used more frequently in the word-medial position than in the word-final position. Three syllable initial consonants(/k, I, s'/) were used more frequently in the word-medial position than in the word-initial position. The syllable and phoneme frequencies in the Korean children's spontaneous speech will provide valuable information in interpreting the severity of phonological disorder and in developing tools for the Korean phonological assessment and intervention.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.30B no.4
• /
• pp.1-10
• /
• 1993

In this paper, a new method of phoneme segmentation of handwritten Korean characters using the local graph pattern is proposed. At first, thinning was performed before extracting features. End-point, inflexion-point, branch-point and cross-point were extracted as features. Using these features and the angular relations between these features, local graph pattern was made. When local graph pattern is made, the of strokes is investigated on contacting point. From this process, pattern is simplified as contacting pattern of the basic form and the contacting form we must take into account can be restricted within fixed region, 4therefore phoneme segmentation not influenced by characters form and any other contact in a single character is performed as matching this local graph pattern with base patterns searched ahead. This experiments with 540 characters have been conducted. From the result of this experiment, it is shown that phoneme segmentation is independent of characters form and other contact in a single character to obtain a correct segmentation rate of 95%, manages it efficiently to reduce the time spent in lock operation when the lock.

• Korean Journal of Child Studies
• /
• v.26 no.4
• /
• pp.145-155
• /
• 2005

This study was carried out to reveal Hangul acquisition processes in Korean-Chinese children who grow in a horizontal bilingual environment. In this experiment Grapheme substitution/deletion tasks and sensible/non-sensible Kulja reading tasks were administered to 3-, 4-, 5- and 6-year-old Korean-Chinese children growing up in a bilingual environment. Results were that Korean-Chinese children showed similar patterns of Hangul acquisition processes to Korean children but acquired grapheme-phoneme(G-P) correspondence earlier than Korean children. Hangul acquisition rates were 41.7%, 45.7%, 53% and 92.7% at age 3, 4, 5 and 6, respectively. Both Korean-Chinese and Korean children showed higher sensitivity for the final consonant than for the initial and middle consonants. Correlation between phoneme perception and reading was only significant among 6-year-olds in non-sensible Kulja reading tasks. Training in transforming ideographic Chinese to a phonetic system could effect early acquisition of G-P correspondence in Korean-Chinese children.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2010.05a
• /
• pp.981-984
• /
• 2010

In this paper, we used Support Vector Machines(SVMs) as the learning method, one of Artificial Neural Network, to segregated from the continuous speech into phonemes, an initial, medial, and final sound, and then, performed continuous speech recognition from it. A Decision boundary of phoneme is determined by algorithm with maximum frequency in a short interval. Speech recognition process is performed by Continuous Hidden Markov Model(CHMM), and we compared it with another phoneme segregated from the eye-measurement. From the simulation results, we confirmed that the method, SVMs, we proposed is more effective in an initial sound than Gaussian Mixture Models(GMMs).

• Journal of the Institute of Electronics and Information Engineers
• /
• v.54 no.5
• /
• pp.85-89
• /
• 2017

In this papers, we apply high performance hardware and machine learning algorithm to build an advanced VAD algorithm for speech enhancement. Since speech is made of series of phoneme, using recurrent neural network (RNN) which consider previous data is proper method to build a speech model. It is impossible to study every noise in real world. So our algorithm is builded by phoneme based study. we detect voice present frames in noisy speech signal and make enhancement of the speech signal. Phoneme based RNN model shows advanced performance in speech signal which has high correlation among each frames. To verify the performance of proposed algorithm, we compare VAD result with label data and speech enhancement result in various noise environments with previous speech enhancement algorithm.

• The Journal of the Acoustical Society of Korea
• /
• v.13 no.1
• /
• pp.5-13
• /
• 1994

This paper presents a study on Neural Networks for Phoneme Recognition and performs the Phoneme Recognition using TDNN (Time Delay Neural Network). Also, this paper proposes training algorithm for speech recognition using neural nets that is a proper to large scale TDNN. Because Phoneme Recognition is indispensable for continuous speech recognition, this paper uses TDNN to get accurate recognition result of phonemes. And this paper proposes new training algorithm that can converge TDNN to an optimal state regardless of the number of phonemes to be recognized. The recognition experiment was performed with new training algorithm for TDNN that combines backpropagation and Cauchy algorithm using stochastic approach. The results of the recognition experiment for three phoneme classes for two speakers show the recognition rates of $98.1\%$. And this paper yielded that the proposed algorithm is an efficient method for higher performance recognition and more reduced convergence time than TDNN.

• The Journal of the Acoustical Society of Korea
• /
• v.19 no.1
• /
• pp.4-10
• /
• 2000

The work presented in this paper is about a postprocessor, which improves the performance of automatic speech segmentation system by correcting the phoneme boundary errors. We propose a postprocessor that reduces the range of errors in the auto labeled results that are ready to be used directly as synthesis unit. Starting from a baseline automatic segmentation system, our proposed postprocessor trains the features of hand labeled results using multi-layer perceptron(MLP) algorithm. Then, the auto labeled result combined with MLP postprocessor determines the new phoneme boundary. The details are as following. First, we select the feature sets of speech, based on the acoustic phonetic knowledge. And then we have adopted the MLP as pattern classifier because of its excellent nonlinear discrimination capability. Moreover, it is easy for MLP to reflect fully the various types of acoustic features appearing at the phoneme boundaries within a short time. At the last procedure, an appropriate feature set analyzed about each phonetic event is applied to our proposed postprocessor to compensate the phoneme boundary error. For phonetically rich sentences data, we have achieved 19.9 % improvement for the frame accuracy, comparing with the performance of plain automatic labeling system. Also, we could reduce the absolute error rate about 28.6%.

• Journal of Broadcast Engineering
• /
• v.25 no.5
• /
• pp.742-749
• /
• 2020

This paper proposes a method for recognizing both phoneme and emotion using a common neural network and a multi-task training method for the common neural network. The common neural network performs the same function for both recognition tasks, which corresponds to the structure of multi-information recognition of human using a single auditory system. The multi-task training conducts a feature modeling that is commonly applicable to multiple information and provides generalized training, which enables to improve the performance by reducing an overfitting occurred in the conventional individual training for each information. A method for increasing phoneme recognition performance is also proposed that applies weight to the phoneme in the multi-task training. When using the same feature vector and neural network, it is confirmed that the proposed common neural network with multi-task training provides higher performance than the individual one trained for each task.

• Journal of Digital Convergence
• /
• v.14 no.6
• /
• pp.499-511
• /
• 2016

The aim of this study was to find out effect of the orthographic representation on speech sound segmentation performance. Children's performances of the orthographic representation task and the speech sound segmentation task had positive correlation in words of phoneme-grapheme correspondence and negative correlation in words of phoneme-grapheme non-correspondence. In the case of words of phoneme-grapheme correspondence, there was no difference in performance ability between orthographic representation high level group and low level group, while in the case of words of phoneme-grapheme non-correspondence, the low level group's performance was significantly better than the high level group's. The most frequent errors of both groups were orthographic conversion errors and such errors were significantly more noticeable in the high level group. This study suggests that from the time of learning orthographic knowledge, children utilize orthographic knowledge for the performance of phonological awareness tasks.

• Journal of Convergence for Information Technology
• /
• v.10 no.8
• /
• pp.35-39
• /
• 2020

DNN error is small compared to the conventional speech recognition system, DNN is difficult to parallel training, often the amount of calculations, and requires a large amount of data obtained. In this paper, we generate a phoneme unit to estimate the GMM parameters with each phoneme model parameters from the GMM to solve the problem efficiently. And it suggests ways to improve performance through clustering for a specific vocabulary to effectively apply them. To this end, using three types of word speech database was to have a DB build vocabulary model, the noise processing to extract feature with Warner filters were used in the speech recognition experiments. Results using the proposed method showed a 97.9% recognition rate in speech recognition. In this paper, additional studies are needed to improve the problems of improved over fitting.