• Journal of Institute of Control, Robotics and Systems
• /
• v.22 no.10
• /
• pp.797-802
• /
• 2016

This paper proposes a noble voice recognition method based on an adaptive MFCC and deep learning for embedded systems. To enhance the recognition ratio of the proposed voice recognizer, ambient noise mixed into the voice signal has to be eliminated. However, noise filtering processes, which may damage voice data, diminishes the recognition ratio. In this paper, a filter has been designed for the frequency range within a voice signal, and imposed weights are used to reduce data deterioration. In addition, a deep learning algorithm, which does not require a database in the recognition algorithm, has been adapted for embedded systems, which inherently require small amounts of memory. The experimental results suggest that the proposed deep learning algorithm and HMM voice recognizer, utilizing the proposed adaptive MFCC algorithm, perform better than conventional MFCC algorithms in its recognition ratio within a noisy environment.

• Phonetics and Speech Sciences
• /
• v.9 no.2
• /
• pp.119-124
• /
• 2017

This study investigates how perception of easy sentences varies with prosody pattern, sound intensity, and signal-to-noise ratio(SNR) in young adults with normal hearing who were in their 20's. The results showed that the presence of proper prosody pattern in the sentences increased correct perception rate of the target sentences, and that the lower the intensity and SNR, the lower the sentence perception scores. The results also showed that SNR had a greater effect on the sentence perception scores than sound intensity. There was a significant decrease of perception scores starting at the level of 15 dB and +3 SNR for the sentences with prosody pattern, while starting at the level of 18 dB and +6 SNR for the sentences without prosody pattern, ending up with a very poor perception score as sound intensity and SNR gets lower. There was a significant difference in the perception score of the sentences with prosody pattern between 20 year-old group and 21 year or older group in several listening conditions of sound intensity and SNR.

• The Proceeding of the Korean Institute of Electromagnetic Engineering and Science
• /
• v.6 no.3
• /
• pp.36-54
• /
• 1995

The error performance of digital radio signals (i.e., M-ary PSK signal, DQPSK signal, MSK signal, GMSK signal) interfered by impulsive noise and electromagnetic interference (EMI) is analyzed and discussed. In analysis at first, the error rate equations have been derived in an electromagnetic interference plus impulsive noise environment. And then, the error performance has been evaluated and shown in figures as a function of carrier-to-noise ratio, carrier-to-interference ratio, impu- lsive index, gaussian noise to impulsive noise power ratio, and interference index to measure the amount of error degradation in digital radio signals. From the obtained results we have known that in the presence of m-distributed tone interference plus inpulsive noise, the more significant the electromagnetic interference amplitude varies, the more significant performance degradation is produced. The listing the digital radio signals from the most degraded to the least is that DQPSK, GMSK, QPSK and MSK signal. In the constant amplitude tone interference plus impulsive noise environment, the effect of in- terference nearly disappears over about 20dB in CIR. The effect of constant tone interference on error rate performance is reduced more remarkably in the region from 10dB to 15dB in CIR. In both enviroments of m-distributed tone interference and constant amplitude tone interference, the more electromagnetic interference amplitude varies and CIR increases, the more error perfor- mance is improved. But it is found out that the performance can not be improved significantly even the electromagnetic interference becomes weak. This describes that the impulsive noise affects dominantly to the performance degradation.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.18 no.4
• /
• pp.949-956
• /
• 2014

Recently, with the development of hardware and software technology related with image information delivery, the demand for various multimedia service has increased. But, the process of treating, sending, and storing image signals generates image degradation by various external causes. The main cause of image degradation is noise. image is mostly damaged by AWGN (additive white Gaussian noise). Therefore, there have been active researches on noise elimination. This paper, to reduce the effects of AWGN added to the image, suggests a noise-eliminating algorithm which is excellent in low-frequency and high-frequency characteristics in space. And, this paper, through simulation techniques, compared the result of the suggested algorithm with those of the existing methods. And, to evaluate the performance of it, PSNR (peak signal to noise ratio) was used.

• Proceedings of the KSME Conference
• /
• 2004.11a
• /
• pp.585-590
• /
• 2004

Many machine failures are not detected well in advance due to the masking of background noise and attenuation of the source signal through the transmission mediums. Advanced signal processing techniques using adaptive filters and higher order statistics have been attempted to extract the source signal from the measured data at the machine surface. In this paper, blind deconvolution using the eigenvector algorithm (EVA) technique is used to recover a damaged bearing signal using only the measured signal at the machine surface. A damaged bearing signal corrupted by noise with varying signal-to-noise (s/n) was used to determine the effectiveness of the technique in detecting an incipient signal and the optimum choice of filter length. The results show that the technique is effective in detecting the source signal with an s/n ratio as low as 0.21, but requires a relatively large filter length.

• Journal of the Korean Magnetic Resonance Society
• /
• v.5 no.2
• /
• pp.66-72
• /
• 2001

The noise suppression of time domain NMR data by discrete wavelet transform with high order Daubechies wavelet coefficients exhibits severe peak distortion and incomplete noise suppression near real signal. However, the fact that even a shift averaged Harr wavelet transform with a set of Daubechies wavelet coefficients (1/2, -l/2) can be used as a new and excellent tool to distinguish real peaks from the noise contaminated NMR signal is introduced. New algorithms of shift averaged Harr wavelet were developed and quantitatively evaluated in terms of threshold and signal to noise ratio (SNR).

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.18 no.10
• /
• pp.2523-2532
• /
• 1994

In this paper, active noise control is performed in a duct system using the periodic pulse train which corresponds to the periodic component of noise source as a reference signal. Control algorithm applied in this study is possible to eliminate the acoustic feedback which occurs in the conventional filtered-x and filtered-u LMS algorithm by using electrical reference signal and has the fast adaptation speed with low filter orders by using synchronous sampling method is discussed via computer simulations and experiments of case studies such as frequency modulation, amplitude modulation and frequency differency between source signal and reference signal.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.38A no.4
• /
• pp.350-359
• /
• 2013

In this paper, Inter user interference elimination algorithm based on Block Diagonal Geometric Mean Decomposition(BD-GMD) for eliminating inter user interference apply to Zero-Forcing in the Successive Signal to Leakage plus Noise Ratio(SSLNR) in Coordinated Multi-Point Coordinated Beamforming system(CoMP CB). As a result, the leakage power is eliminated. The inter user interference elimination algorithm, however, cannot guarantee the enough desired signal power therefore we perform the channel ordering to overcome this disadvantage and increase the desired signal power. The simulation results show that the proposed scheme provides the improved Bit Error Rate(BER) performance compared with existing SSLNR-Zero-Forcing-Tomlinson Harashima precoding(SSLNR-ZF-THP).

• Proceedings of the KSRS Conference
• /
• 2004.10a
• /
• pp.444-447
• /
• 2004

SNR(Signal to Noise Ratio) is one of the most important performance for the electro-optical camera system. This paper shows not only the SNR analyses for the MSC system, which is the payload in the KOMPSAT2 satellite, but also the trials for its improvement in the electronics circuit design. The MSC deals with one panchromatic band and four multi-spectral bands. The SNR analyses are performed based on the MSC design for the each band and assuming that the defined radiance reached directly to the sensor entrance pupil. In the SNR calculation, shot noise, dark current noise, analog electronics noise and ADC quantization noise are considered as noise sources. In these noise sources, especially, the electronics noise depends on the camera electronics design. This paper shows the camera electronics design to increase SNR and its test results as well as the SNR analyses.

• Journal of Electrical Engineering and Technology
• /
• v.11 no.6
• /
• pp.1793-1801
• /
• 2016

Speech enhancement is a challenging problem due to the diversity of noise sources and their effects in different applications. The goal of speech enhancement is to improve the quality and intelligibility of speech by reducing noise. Many research works in speech enhancement have been accomplished in English and other European Languages. There has been limited or no such works or efforts in the past in the context of Tamil speech enhancement in the literature. The aim of the proposed method is to reduce the background noise present in the Tamil speech signal by using wavelets. New modified thresholding function is introduced. The proposed method is evaluated on several speakers and under various noise conditions including White Gaussian noise, Babble noise and Car noise. The Signal to Noise Ratio (SNR), Mean Square Error (MSE) and Mean Opinion Score (MOS) results show that the proposed thresholding function improves the speech enhancement compared to the conventional hard and soft thresholding methods.