• Phonetics and Speech Sciences
• /
• v.12 no.4
• /
• pp.63-71
• /
• 2020

Speech recognition is one of the areas actively commercialized using deep learning and machine learning techniques. However, the majority of speech recognition systems on the market are developed on data with limited diversity of speakers and tend to perform well on typical adult speakers only. This is because most of the speech recognition models are generally learned using a speech database obtained from adult males and females. This tends to cause problems in recognizing the speech of the elderly, children and people with dialects well. To solve these problems, it may be necessary to retain big database or to collect a data for applying a speaker adaptation. However, this paper proposes that a new end-to-end speech recognition method consists of an acoustic augmented recurrent encoder and a transformer decoder with linguistic prediction. The proposed method can bring about the reliable performance of acoustic and language models in limited data conditions. The proposed method was evaluated to recognize Korean elderly and children speech with limited amount of training data and showed the better performance compared of a conventional method.

• Proceedings of the IEEK Conference
• /
• 2002.07a
• /
• pp.690-693
• /
• 2002

According to the growth of communication biz, embedded market rapidly developing in domestic and overseas. Embedded system can be used in various way such as wire and wireless communication equipment or information products. There are lots of developing performance applying speech recognition to embedded system, for instance, PDA, PCS, CDMA-2000 or IMT-2000. This study implement minimum memory of speech recognition engine and DB for apply real time embedded system. The implement measure of speech recognition equipment to fit on embedded system is like following. At first, DC element is removed from Input voice and then a compensation of high frequency was achieved by pre-emphasis with coefficients value, 0.97 and constitute division data as same size as 256 sample by lapped shift method. Through by Levinson - Durbin Algorithm, these data can get linear predictive coefficient and again, using Cepstrum - Transformer attain feature vectors. During HMM training, We used Baum-Welch reestimation Algorithm for each words training and can get the recognition result from executed likelihood method on each words. The used speech data is using 40 speech command data and 10 digits extracted form each 15 of male and female speaker spoken menu control command of Embedded system. Since, in many times, ARM CPU is adopted in embedded system, it's peformed porting the speech recognition engine on ARM core evaluation board. And do the recognition test with select set 1 and set 3 parameter that has good recognition rate on commander and no digit after the several tests using by 5 proposal recognition parameter sets. The recognition engine of recognition rate shows 95%, speech commander recognizer shows 96% and digits recognizer shows 94%.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.31B no.1
• /
• pp.11-22
• /
• 1994

For the conventional systme based on word, it is very difficult to enlarge the number of vocabulary. To cope with this problem, we must use more fundamental units of speech. For example, syllables and phonemes are such units, Korean speech consists of initial consonants, middle vowels and final consonants and has characteristic that we can obtain syllables from speech easily. In this paper, we show a speech recognition system with the advantage of the syllable characteristics peculiar to the Korean speech. The algorithm of recognition system is the Time Delay Neural Network. To recognize many recognition units, system consists of initial consonants, middle vowels, and final consonants recognition neural network. At first, our system recognizes initial consonants, middle vowels and final consonants. Then using this results, system recognizes isolated words. Through experiments, we got 85.12% recognition rate for 2735 data of initial consonants, 86.95% recognition rate for 3110 data of middle vowels, and 90.58% recognition rate for 1615 data of final consonants. And we got 71.2% recognition rate for 250 data of isolated words.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.8 no.2
• /
• pp.105-110
• /
• 2008

Human being recognizes emotion fusing information of the other speech signal, expression, gesture and bio-signal. Computer needs technologies that being recognized as human do using combined information. In this paper, we recognized five emotions (normal, happiness, anger, surprise, sadness) through speech signal and facial image, and we propose to method that fusing into emotion for emotion recognition result is applying to multimodal method. Speech signal and facial image does emotion recognition using Principal Component Analysis (PCA) method. And multimodal is fusing into emotion result applying fuzzy membership function. With our experiments, our average emotion recognition rate was 63% by using speech signals, and was 53.4% by using facial images. That is, we know that speech signal offers a better emotion recognition rate than the facial image. We proposed decision fusion method using S-type membership function to heighten the emotion recognition rate. Result of emotion recognition through proposed method, average recognized rate is 70.4%. We could know that decision fusion method offers a better emotion recognition rate than the facial image or speech signal.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.14 no.2
• /
• pp.201-206
• /
• 2004

This study attempted to develop a speech recognition module applied to a wheelchair for the physically handicapped. In the proposed speech recognition module, TMS320C32 was used as a main processor and Mel-Cepstrum 12 Order was applied to the pro-processor step to increase the recognition rate in a noisy environment. DTW (Dynamic Time Warping) was used and proven to be excellent output for the speaker-dependent recognition part. In order to utilize this algorithm more effectively, the reference data was compressed to 1/12 using vector quantization so as to decrease memory. In this paper, the necessary diverse technology (End-point detection, DMA processing, etc.) was managed so as to utilize the speech recognition system in real time

• Journal of the Institute of Electronics Engineers of Korea TE
• /
• v.37 no.3
• /
• pp.16-20
• /
• 2000

Most of speech recognition systems are using Hidden Markov Model based on statistical modelling frequently. In Korean isolated telephone digit speech recognition, high recognition rate is gained by using HMM if many training data are given. But in Korean continuous telephone digit speech recognition, HMM has some limitations for similar telephone digits. In this paper we suggest a way to overcome some limitations of HMM by using discriminant learning based on minimal classification error criterion in Korean continuous telephone digit speech recognition. The experimental results show our method has high recognition rate for similar telephone digits.

• Journal of Biomedical Engineering Research
• /
• v.27 no.3
• /
• pp.101-109
• /
• 2006

It has been known that there is strong relationship between human voices and the movements of the articulatory facial muscles. In this paper, we utilize this knowledge to implement an automatic speech recognition scheme which uses solely surface electromyogram (EMG) signals. The EMG signals were acquired from three articulatory facial muscles. Preliminary, 10 Korean digits were used as recognition variables. The various feature parameters including filter bank outputs, linear predictive coefficients and cepstrum coefficients were evaluated to find the appropriate parameters for EMG-based speech recognition. The sequence of the EMG signals for each word is modelled by a hidden Markov model (HMM) framework. A continuous word recognition approach was investigated in this work. Hence, the model for each word is obtained by concatenating the subword models and the embedded re-estimation techniques were employed in the training stage. The findings indicate that such a system may have a capacity to recognize speech signals with an accuracy of up to 90%, in case when mel-filter bank output was used as the feature parameters for recognition.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.8 no.3
• /
• pp.64-70
• /
• 2009

This paper presents a speech enhancement technique based on generalized Gamma distribution in order to obtain robust speech recognition performance. For robust speech enhancement, the noise estimation based on a spectral noise floor controled recursive averaging spectral values is applied to speech estimation under the generalized Gamma distribution and spectral gain floor. The proposed speech enhancement technique is based on spectral component, spectral amplitude, and log spectral amplitude. The performance of three different methods is measured by recognition accuracy of automatic speech recognition (ASR).

• The Journal of the Acoustical Society of Korea
• /
• v.29 no.2E
• /
• pp.86-99
• /
• 2010

This paper analyzes the performance of various single channel speech enhancement algorithms when they are applied to automatic speech recognition (ASR) systems as a preprocessor. The functional modules of speech enhancement systems are first divided into four major modules such as a gain estimator, a noise power spectrum estimator, a priori signal to noise ratio (SNR) estimator, and a speech absence probability (SAP) estimator. We investigate the relationship between speech recognition accuracy and the roles of each module. Simulation results show that the Wiener filter outperforms other gain functions such as minimum mean square error-short time spectral amplitude (MMSE-STSA) and minimum mean square error-log spectral amplitude (MMSE-LSA) estimators when a perfect noise estimator is applied. When the performance of the noise estimator degrades, however, MMSE methods including the decision directed module to estimate a priori SNR and the SAP estimation module helps to improve the performance of the enhancement algorithm for speech recognition systems.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.5 no.6
• /
• pp.568-574
• /
• 2010

In an effort to improve the speech recognition rate, we investigated performance comparison between speaker-independent and gender-specific speech recognitions. For this purpose, 20 male and 20 female speakers each pronounced 300 isolated Korean words and the speeches were divided into 4 groups: female, male, and two mixed genders. To examine the validity for the gender-specific speech recognition, Fourier spectrum and MFCC feature vectors averaged over male and female speakers separately were examined. The result showed distinction between the two genders, which supports the motivation for the gender-specific speech recognition. In experiments of speech recognition rate, the error rate for the gender-specific case was shown to be less than50% compared to that of the speaker-independent case. From the obtained results, it might be suggested that hierarchical recognition of gender and speech recognition might yield better performance over the current method of speech recognition.