• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.3 no.1
• /
• pp.185-196
• /
• 1999

In this dissertation, a new off-line recognition system is proposed using a subgraph pattern, neural network. After thinning is applied to input characters, balance having a noise elimination function on location is performed. Then as the first step for recognition procedure, circular elements are extracted and recognized. From the subblock HT, space feature points such as endpoint, flex point, bridge point are extracted and a subgraph pattern is formed observing the relations among them. A region where vowel can exist is allocated and a candidate point of the vowel is extracted. Then, using the subgraph pattern dictionary, a vowel is recognized. A same method is applied to extract horizontal vowels and the vowel is recognized through a simple structural analysis. For verification of recognition subgraph in this paper, experiments are done with the most frequently used Myngjo font, Gothic font for printed characters and handwritten characters. In case of Gothic font, character recognition rate was 98.9%. For Myngjo font characters, the recognition rate was 98.2%. For handwritten characters, the recognition rate was 92.5%. The total recognition rate was 94.8% with mixed handwriting and printing characters for multi-font recognition.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.19 no.6
• /
• pp.1140-1148
• /
• 1994

In this paper, we carried out some experiments on the Korean vowel recognition using the fractal dimension of the speech signals. We chose the Minkowski-Bouligand dimension as the fractal dimension, and computed it using the morphological covering method. For our experiments, we used both the fractal dimension and the LPC cepstrum which is conventionally known to be one of the best parameters for speech recognition, and examined the usefulness of the fractal dimension. From the vowel recognition experiments under various consonant contexts, we achieved the vowel recognition error rates of 5.6% and 3.2% for the case with only LPC cepstrum and that with both LPC cepstrum and the fractal dimension, respectively. The results indicate that the incorporation of the fractal dimension with LPC cepstrum gives more than 40% reduction in recognition errors, and indicates that the fractal dimension is a useful feature parameter for speech recognition.

• The Journal of the Acoustical Society of Korea
• /
• v.11 no.1
• /
• pp.35-41
• /
• 1992

A speaker identification scheme using the speaker-based VQ codecook of a sustained vowel is proposed and tested. With the pitch synchronous LPC vector of the sustained vowel /i/ as a feature vector, a VQ codebook size of 4 was found to be suitable to characterize each speaker's feature space. For 40 normal speakers (20 males, 20 females), we achieved the correct identification rate of 99.4% with a training data set, and 89.4% with a test data set with speech samples of only 50 pitch periods.

• Annual Conference on Human and Language Technology
• /
• 1997.10a
• /
• pp.194-199
• /
• 1997

본 논문에서는 글꼴의 변화와 잡영을 흡수할 수 있도록 자소의 탐색 영역을 정의 하였으며 이 영역에 나타나는 횡모음과 종모음의 주획을 추출하는 방법을 기술하였다. 종모음 영역에서 추출한 수직획들과 횡모음 영역에서 추출한 수평획들을 각각 종모음과 횡모음의 주획이 될 수 있는 후보들로써 이들로 부터 종모음과 횡모음의 존재를 파악하는 것이 한글 유형 분류의 주된 내용이다. 그러나 다양한 글꼴에 나타나는 수평획들로부터 곧바로 횡모음의 존재를 파악하는 것은 쉬운 문제가 아니다 본 논문에서는 기존의 트리 분류기를 확장하여 복잡하고 다양한 특징을 단계별로 단순화시키고 트리 분류기의 상위 노드에서 결정된 정보와 제약 조건을 이용하여 유형을 분류하는 방법을 제안하였다. 제안된 방법은 한글 상위 빈도 1405자, 3가지 글꼴에 대하여 99.8 %의 유형 분류율을 보이고 있다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2011.10a
• /
• pp.613-615
• /
• 2011

본 논문에서는 켑스트럼 계수를 이용하여 음성인식을 하는 알고리즘을 제안한다. 본 논문에서 제안하는 방법은 사람이 발성한 음성을 두 영역의 켑스트럼 계수로 분리한 후에, 신경회로망을 사용하여 음성인식을 하는 방법이다. 본 논문에서 제안하는 신경회로망은 오차가 거의 없어지는 일정 기간 동안 네트워크를 학습시킨 후에 신경회로망의 학습 데이터와는 다른 새로운 음성이 신경회로망에 입력된 경우에 대하여 각 음성 구간에서 분류가 가능한 모음검출을 위한 음성인식 시스템을 제안한다.

• KIISE Transactions on Computing Practices
• /
• v.22 no.11
• /
• pp.583-590
• /
• 2016

Recently, the rapidly increasing interest in IoT in almost all areas of casual human life has led to wide acceptance of speech recognition as a means of HCI. Simultaneously, the demand for speech recognition systems for mobile environments is increasing rapidly. The server-based speech recognition systems are typically fast and show high recognition rates; however, an internet connection is necessary, and complicated server computation is required since a voice is recognized by units of words that are stored in server databases. In this paper, we present a novel method for recognizing the Korean vowel 'ㅗ', as a part of a phoneme based Korean speech recognition system. The proposed method involves analyses of waveform patterns in the time domain instead of the frequency domain, with consequent reduction in computational cost. Elementary algorithms for detecting typical waveform patterns of 'ㅗ' are presented and combined to make final decisions. The experimental results show that the proposed method can achieve 89.9% recognition accuracy.

• KIISE Transactions on Computing Practices
• /
• v.20 no.11
• /
• pp.567-572
• /
• 2014

In recent years, computing and networking technologies have been developed, and the communication equipments have become smaller and the mobility has increased. In addition, the demand for easily-operated speech recognition has increased. This paper proposes method of recognizing the Korean phoneme 'ㅏ'. A phoneme is the smallest unit of sound, and it plays a significant role in speech recognition. However, the precise recognition of the phonemes has many obstacles since it has many variations in its pronunciation. This paper proposes a simple and efficient method that can be used to recognize a Korean vowel 'ㅏ'. The proposed method is based on the common features that are extracted from the 'ㅏ' waveform sequences, and this is simpler than when using the previous complex methods. The experimental results indicate that this method has a more than 90 percent accuracy in recognizing 'ㅏ'.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2001.10a
• /
• pp.789-792
• /
• 2001

본 논문은 자동차의 번호판 인식 시스템의 한 부분인 번호판 추출과 자모 분리를 통한 문자 인식까지의 과정을 실험한 것이다. 본 논문은 gray-level에서 영상을 실험하였고, 번호판을 추출하기 위해서 morphology를 반복 적용하고 크기 보정을 통해 번호판을 추출하며, hough transform을 이용한 크기 재보정을 통해 최종적으로 번호판을 추출한다. 그리고, 문자 인식 단계에서는 먼저 hough transform을 사용하여 한글의 모음의 시작점을 얻고, 문자 특징을 이용하여 자음과 모음을 분리하여 모음을 인식한다.

• MALSORI
• /
• no.48
• /
• pp.69-80
• /
• 2003

음성 영어교육에서 각 음소(모음.자음) 음가 인식이 중요한 것은 물론이지만, 실제로 영어를 듣고 말하는 데는 강세모음과 약음(주로 음)이 번갈아들며 엮어내는 영어 특유 말씨 박자에 대한 이해가 오히려 더 중요한 면이 있다. 그런데 영어 철자가 자음 정보는 비교적 잘 보여주지만, 모음 정보는 정확히 알려주지 못한다. 이를테면 (equation omitted)를 적는 철자가 각각 20가지 이상이며, 특히 쭉정모음(schwa) $\partial$를 적는 철자는 무려 90가지가 넘는다. 이 $\partial$는 강세 그늘에서 비록 짧고 약하게 발음되지만 영어에서 딴 어느 모음보다도 훨씬 높은 빈도를 보이기 때문에 [(equation omitted)] 음가와 그것이 있는 자리, 즉 철자 밑에 숨어 그 존재를 알기 어려운 $\partial$ 음의 소재를 철저히 파악하는 것은 영어를 잘 듣고 영어답게 말하는 데에 필수 요체라 할 수 있다. 이 글에서는 쭉정모음(schwa) [$\partial$]를 포함한 영어 주요 모음 철자 다양성을 고찰하여 우리나라 음성 영어 교육(TEFL/TESOL)에 귀중한 참고 자료를 제공하고, 이를 통한 올바른 영어 발음 학습 태도를 제언한다.

• The Journal of the Acoustical Society of Korea
• /
• v.32 no.2
• /
• pp.167-173
• /
• 2013

This work examines classification of diphthongs, as part of a distinctive feature-based speech recognition system. Acoustic measurements related to the vocal tract and the voice source are examined, and analysis of variance (ANOVA) results show that vowel duration, energy trajectory, and formant variation are significant. A balanced error rate of 17.8% is obtained for 2-way diphthong classification on the TIMIT database, and error rates of 32.9%, 29.9%, and 20.2% are obtained for /aw/, /ay/, and /oy/, for 4-way classification, respectively. Adding the acoustic features to widely used Mel-frequency cepstral coefficients also improves classification.