• Proceedings of the KSPS conference
• /
• 1996.02a
• /
• pp.137-155
• /
• 1996

The purpose of this paper is to generate a pitch contour which is affected by tile phonetic environment and the number of syllables in each Korean isolated word using a neural network. To do this, we analyzed a set of 513 Korean isolated words, consisting of 1-4 syllables and extracted the pitch contour and the duration of each phoneme in all the words. The total number of phonemes we analyzed is about 3800. After that we approximated the pitch contour with a 1st order polynominal by a regression analysis. We could get the slope, the initial pitch and the duration of each phoneme. We used these 3 parameters as the target pattern of the neural network and let the neural network learn the rule of the variation of the pitch and duration, which was affected by the phonetic environment of each phoneme. We used 7 consecutive phoneme strings as an input pattern for a neural network to make the network learn the effect of phonetic environment around the center phoneme. In the learning phase, we used 3545 items(463 words) as target patterns which contained the phonetic environment of front and rear 3 phonemes and the neural network showed the correctness rate of 98.43%, 98.59%, 97.7% in the estimation of the duration, the slope, the initial pitch. In the recall phase, we tested the performance of tile neural network with 251 items(50 words) which weren't need as learning data and we could get the good correctness rate of 97.34%, 95.45%, 96.3% in the generation of the duration, the slope, and the initial pitch of each phoneme.

• The Journal of the Acoustical Society of Korea
• /
• v.14 no.3
• /
• pp.49-60
• /
• 1995

This paper is a study on the design and implementation of the Korean Text-to-Speech system which is used for a small and simple system. In this paper, a parameter synthesis method is chosen for speech syntheiss method, we use PARCOR(PARtial autoCORrelation) coefficient which is one of the LPC analysis. And we use phoneme for synthesis unit which is the basic unit for speech synthesis. We use PARCOR, pitch, amplitude as synthesis parameter of voice, we use residual signal, PARCOR coefficients as synthesis parameter of unvoice. In this paper, we could obtain the 60% intelligibility by using the residual signal as excitation signal of unvoiced sound. The result of synthesis experiment, synthesis of a word unit is available. The controlling of phoneme duration is necessary for synthesizing of a sentence unit. For setting up the synthesis system, PC 486, a 70[Hz]-4.5[KHz] band pass filter for speech input/output, amplifier, and TMS320C30 DSP board was used.

• Journal of the Korean BIBLIA Society for library and Information Science
• /
• v.21 no.1
• /
• pp.131-148
• /
• 2010

This study aimed to develop an operating manual on the voluntary activity to the library services with disabled people in the public libraries. Various domestic and overseas manuals on the voluntary activity to the library services with disabled people were analyzed. From the survey and expert interviews, five types of most frequently asked library services with disabled people, face-to-face reading, word input services, services to off-library users, mobility aids services and assistance of engineering devices were extraced. Five specific areas on the voluntary activity were chosen in terms of activity area, barrier type, specific activity field, activity purpose and checklist. Finally, contents of the operating manual was applied to the area of face-to-face reading.

• Journal of Intelligence and Information Systems
• /
• v.10 no.2
• /
• pp.65-77
• /
• 2004

This paper presents a priori and the local font classification method. The font classification uses ascenders, descenders, and serifs extracted from a word image. The gradient features of those sub-images are extracted, and used as an input to a neural network classifier to produce font classification results. The font classification determines 2 font styles (upright or slant), 3 font groups (serif sans-serif or typewriter), and 7-font names (Postscript fonts such as Avant Garde, Helvetica, Bookman, New Century Schoolbook, Palatine, Times, and Courier). The proposed a priori and local font classification method allows an OCR system consisting of various font-specific character segmentation tools and various mono-font character recognizers. Experiments have shown font classification accuracies reach high performance levels of about 95.4 percent even with severely touching characters. The technique developed for tile selected 7 fonts in this paper can be applied to any other fonts.

• The Journal of the Acoustical Society of Korea
• /
• v.20 no.4
• /
• pp.48-53
• /
• 2001

In this paper, selectively attentive learning method has been proposed to improve the learning speed of multilayer Perceptron based on the error backpropagation algorithm. Three attention criterions are introduced to effectively determine which set of input patterns is or which portion of network is attended to for effective learning. Such criterions are based on the mean square error function of the output layer and class-selective relevance of the hidden nodes. The acceleration of learning time is achieved by lowering the computational cost per iteration. Effectiveness of the proposed method is demonstrated in a speaker adaptation task of isolated word recognition system. The experimental results show that the proposed selective attention technique can reduce the learning time more than 60% in an average sense.

• The KIPS Transactions:PartB
• /
• v.14B no.7
• /
• pp.559-566
• /
• 2007

Subjective sentences express opinions, emotions, evaluations and other subjective ideas relevant to products or events. These expressions sometimes can be seen in only part of a sentence, thus extracting features from a full-sentence can degrade the performance of subjective-sentence-classification. This paper presents a method for improving the performance of a subjectivity classifier by combining two classifiers generated from the different representations of an input sentence. One representation is a sentimental phrase that represents an automatically identified subjective expression or objective expression and the other representation is a full-sentence. Each representation is used to extract modified n-grams that are composed of a word and its contextual words' polarity information. The best performance, 79.7% accuracy, 2.5% improvement, was obtained when the phrase-level classifier and the sentence-level classifier were merged.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.21 no.1
• /
• pp.223-228
• /
• 2017

In this paper, we are proposing an optimal design of wideband pulsed type gamma-ray sensors. These sensors were manufactured based on the design results and after word electrical properties were analyzed. The sensor input parameters were derived on the basis of pulsed gamma-ray spectrum and time-dependent energy rate, and the output current which were derived on the basis of the sensor sensitivity control circuit. Pulsed gamma-ray sensors were designed using the TCAD simulators. The design results show that the optimal Epi layer thickness is 45um with the applied voltage 3.3V and the diameter is 2.0mm. The doping concentrations are as follows : N-type is an Arsenic as $1{\times}10^{19}/cm^3$, P-type is a Boron as $1{\times}10^{19}/cm^3$ and Epi layer is Phosphorus as $3.4{\times}10^{12}/cm^3$. The fabricated sensor was a leakage current, 12pA at voltage -3.3V and fully depleted mode at voltage -5V. A test result of pulsed radiation shows that the sensor gives out the optimal photocurrent.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.48 no.5
• /
• pp.45-51
• /
• 2011

In this paper, we propose an unsupervised learning method for modeling motion trajectory patterns effectively. In our approach, observations of an object on a trajectory are treated as words in a document for latent dirichlet allocation algorithm which is used for clustering words on the topic in natural language process. This allows clustering topics (e.g. go straight, turn left, turn right) effectively in complex scenes, such as crossroads. After this procedure, we learn patterns of word sequences in each cluster using Baum-Welch algorithm used to find the unknown parameters in a hidden markov model. Evaluation of abnormality can be done using forward algorithm by comparing learned sequence and input sequence. Results of experiments show that modeling of semantic region is robust against noise in various scene.

• The Journal of the Acoustical Society of Korea
• /
• v.16 no.5
• /
• pp.50-59
• /
• 1997

In this paper, we implement an automatic speech segmentation and labeling system which marks phone boundaries automatically for constructing the Korean speech database. We specify and implement the system based on conventional speech segmentation and labeling techniques, and also develop the graphic user interface(GUI) on Hangul $Motif^{TM}$ environment for the users to examine the automatic alignment boundaries and to refine them easily. The developed system is applied to 16kHz sampled speech, and the labeling unit is composed of 46 phoneme-like units(PLUs) and silence. The system uses both of the phonetic and orthographic transcription as input methods of linguistic information. For pattern-matching method, hidden Markov models(HMM) is employed. Each phoneme model is trained using the manually segmented 445 phonetically balanced word (PBW) database. In order to evaluate the performance of the system, we test it using another database consisting of sentence-type speech. According to our experiment, 74.7% of phoneme boundaries are within 20ms of the true boundary and 92.8% are within 40ms.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.41 no.5
• /
• pp.592-601
• /
• 2016

The recurrent neural network (RNN) is a deep learning model which is suitable to sequential or length-variable data. The Long Short-Term Memory (LSTM) mitigates the vanishing gradient problem of RNNs so that LSTM can maintain the long-term dependency among the constituents of the given input sequence. In this paper, we propose a LSTM based language model which can predict following words of a given incomplete sentence to generate a complete sentence. To evaluate our method, we trained our model using multiple Korean corpora then generated the incomplete part of Korean sentences. The result shows that our language model was able to generate the fluent Korean sentences. We also show that the word based model generated better sentences compared to the other settings.