• Phonetics and Speech Sciences
• /
• v.15 no.2
• /
• pp.43-51
• /
• 2023

In this paper, we propose an approach for dialect classification based on the speed and pause of speech utterances as well as the age and gender of the speakers. Dialect classification is one of the important techniques for speech analysis. For example, an accurate dialect classification model can potentially improve the performance of speaker or speech recognition. According to previous studies, research based on deep learning using Mel-Frequency Cepstral Coefficients (MFCC) features has been the dominant approach. We focus on the acoustic differences between regions and conduct dialect classification based on the extracted features derived from the differences. In this paper, we propose an approach of extracting underexplored additional features, namely the speed and the pauses of speech utterances along with the metadata including the age and the gender of the speakers. Experimental results show that our proposed approach results in higher accuracy, especially with the speech rate feature, compared to the method only using the MFCC features. The accuracy improved from 91.02% to 97.02% compared to the previous method that only used MFCC features, by incorporating all the proposed features in this paper.

• Proceedings of the IEEK Conference
• /
• 2001.06e
• /
• pp.177-180
• /
• 2001

A saw-tooth waveform generator with current-controllable frequency is described. The generator utilizes operational transconductance amplifiers as switching element. It features simple and wide sweep capability. The circuit built with commercially avaliable components exhibits good linearity of current to frequency and relatively low temperature sensitivity.

• Journal of Broadcast Engineering
• /
• v.18 no.5
• /
• pp.771-774
• /
• 2013

This study investigates the relationship between voice quality measurements and emotional states, in addition to conventional prosodic and cepstral features. Open quotient, harmonics-to-noise ratio, spectral tilt, spectral sharpness, and band energy were analyzed as voice quality features, and prosodic features related to fundamental frequency and energy are also examined. ANOVA tests and Sequential Forward Selection are used to evaluate significance and verify performance. Classification experiments show that using the proposed features increases overall accuracy, and in particular, errors between happy and angry decrease. Results also show that adding voice quality features to conventional cepstral features leads to increase in performance.

• The Journal of the Acoustical Society of Korea
• /
• v.17 no.2E
• /
• pp.31-37
• /
• 1998

This paper proposes a feature extraction method using wavelet transform for speech recognition. Speech recognition system generally carries out the recognition task based on speech features which are usually obtained via time-frequency representations such as Short-Time Fourier Transform (STFT) and Linear Predictive Coding(LPC). In some respects these methods may not be suitable for representing highly complex speech characteristics. They map the speech features with same may not frequency resolutions at all frequencies. Wavelet transform overcomes some of these limitations. Wavelet transform captures signal with fine time resolutions at high frequencies and fine frequency resolutions at low frequencies, which may present a significant advantage when analyzing highly localized speech events. Based on this motivation, this paper investigates the effectiveness of wavelet transform for feature extraction of wavelet transform for feature extraction focused on enhancing speech recognition. The proposed method is implemented using Sampled Continuous Wavelet Transform (SCWT) and its performance is tested on a speaker-independent isolated word recognizer that discerns 50 Korean words. In particular, the effect of mother wavelet employed and number of voices per octave on the performance of proposed method is investigated. Also the influence on the size of mother wavelet on the performance of proposed method is discussed. Throughout the experiments, the performance of proposed method is discussed. Throughout the experiments, the performance of proposed method is compared with the most prevalent conventional method, MFCC (Mel0frequency Cepstral Coefficient). The experiments show that the recognition performance of the proposed method is better than that of MFCC. But the improvement is marginal while, due to the dimensionality increase, the computational loads of proposed method is substantially greater than that of MFCC.

• Journal of Mechanical Science and Technology
• /
• v.18 no.9
• /
• pp.1537-1548
• /
• 2004

This paper presents a novel diagnostic technique for monitoring the system conditions and detecting failure modes and precursors based on wavelet-packet analysis of external noise/vibration measurements. The capability is based on extracting relevant features of noise/vibration data that best discriminate systems with different noise/vibration signatures by analyzing external measurements of noise/vibration in the time-frequency domain. By virtue of their localized nature both in time and frequency, the identified features help to reveal faults at the level of components in a mechanical system in addition to the existence of certain faults. A prima-facie case is made via application of the proposed approach to fault detection in scroll and rotary compressors, although the methods and algorithms are very general in nature. The proposed technique has successfully identified the existence of specific faults in the scroll and rotary compressors. In addition, its capability of tracking the severity of specific faults in the rotary compressors indicates that the technique has a potential to be used as a prognostic tool.

• Journal of KIISE:Software and Applications
• /
• v.30 no.1_2
• /
• pp.31-39
• /
• 2003

It is often assumed that text regions in images are characterized by some distinctive or characteristic spatial frequencies. This feature is highly intuitive, and thus appealing as much. We propose a method of detecting horizontal texts in natural scene images. It is based on the use of two features that can be employed separately or in succession: the frequency of edge pixels across vertical and horizontal scan lines, and the fundamental frequency in the Fourier domain. We confirmed that the frequency features are language independent. Also addressed is the detection of quadrilaterals or approximate rectangles using Hough transform. Since texts that is meaningful to many viewers usually appear within rectangles with colors in high contrast to the background. Hence it is natural to assume the detection rectangles may be helpful for locating desired texts correctly in natural outdoor scene images.

• Journal of the Korean Society for Railway
• /
• v.2 no.2
• /
• pp.21-30
• /
• 1999

This study proposes the analysis and evaluation method of time series ultrasonic signal using the phase space-frequency domain. Features extracted from time series signal analyze quantitatively characteristics of weld defects. For this purpose, analysis objectives in this study are features of time domain and frequency domain. Trajectory changes in the attractor indicated a substantial difference in fractal characteristics resulting from distance shifts such as parts of head and flange even though the types of defects are identified. These differences in characteristics of weld defects enables the evaluation of unique characteristics of defects in the weld zone. In quantitative fractal feature extraction, feature values of 3.848 in the case of part of head(crack) and 4.102 in the case of part of web(side hole) and 3.711 in the case of part of flange(crack) were proposed on the basis of fractal dimension. Proposed phase space-frequency domain method in this study can integrity evaluation for defect signals of rail weld zone such as side hole and crack.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2011.05a
• /
• pp.340-343
• /
• 2011

To distinguish between a group of speech signals and nonspeech signals, you can use several features in domains like frequency, quefrency and time. It is very important to use features that differentiate two signal groups. As a feature to separate two signal groups, autocorrelation method was proposed and the variances between groups were studied. Autocovariances were just calculated for the time domain signal. Signals were divided into segments which consist of 128 data to be transformed to the frequency and quefrency domains. Autocovariances between each coefficient of segments in FFTs and quefrencies were found and they were averaged over wide spectrum. It is clear that the autocovariances in frequency domain show great differences between groups of signals.

• The Research Journal of the Costume Culture
• /
• v.10 no.2
• /
• pp.186-199
• /
• 2002

This study is carried out to examine the purpose of utilization and the degree of satisfaction on the information collected from the Internet site for the understanding of the reality of clothing information for college students. The results are as fellows; 1 . Three primary factors have been abstracted for the satisfaction of information. 2. The difference in the purpose of utilizing information, depending on the individual characteristic of students and the features related to computer, was shown in the difference of major and school year, number of the internet connection per day, capability of using computer, period of using computer, and frequency of information searching. The more the number of internet connection per day, the less the capability of using computer, and searching for information when required, internet can be utilized for study and work. 3. The difference in the degree of information satisfaction, depending on the individual characteristic of students and the features related to computer, was shown in the difference of major, school year, opportunity to take computer-related course, frequency of searching information, and etc. More positive and satisfactory response was derived from these not majoring in clothing rather than those majoring, those taking computer course related to clothing, and those searching for clothing information periodically. 4. The primary factors affecting the satisfaction on the overall information of clothing show difference in terms of sex and frequency of searching information. Female students show lower degree of satisfaction than male students and more satisfaction on the information searched as the frequency of searching becomes less.

• The KIPS Transactions:PartD
• /
• v.13D no.2 s.105
• /
• pp.279-286
• /
• 2006

One of the widely used methods for product recommendation in Internet storefronts is matching product features against target customer profiles. When using this method, it's very important to choose a suitable subset of features for recommendation efficiency and performance, which, however, has not been rigorously researched so far. In this paper, we utilize a dataset collected from a virtual shopping experiment in a Korean Internet book shopping mall to compare several popular methods from other disciplines for selecting features for product recommendation: the vector-space model, TFIDF(Term Frequency-Inverse Document Frequency), the mutual information method, and the singular value decomposition(SVD). The application of SVD showed the best performance in the analysis results.