• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.18 no.7
• /
• pp.1740-1748
• /
• 2014

This paper focuses on listening test and acoustic analysis of patients' speech with velopharyngeal insufficiency (VPI) and normal speakers' simulation speech. In this research, a set consisting of 50-words, vowels and single syllables is determined for speech database construction. A web-based listening evaluation system is developed for a convenient/automated evaluation procedure. The analysis results show the trend of incorrect recognition for VPI speech and the one for simulation speech are similar. Such similarity is also confirmed by comparing the formant locations of vowel and spectrum of consonant sounds. These results show that the simulation method for VPI speech is effective at generating the speech signals similar to actual VPI patient's speech. It is expected that the simulation speech data can be effectively employed for our future work such as acoustic model adaptation.

• Phonetics and Speech Sciences
• /
• v.2 no.3
• /
• pp.47-63
• /
• 2010

This study provides a description of the glottal characteristics of the word-initial low vowels /a, $\ae$/ in terms of a set of acoustic parameters and discusses glottal configuration as their acoustic correlates. Furthermore, it examines the effect of prosodic boundary on the glottal properties of the vowels, seeking an account of the possible role of prosodic structure based on prosodic theory. Acoustic parameters reported to indicate glottal characteristics were obtained from the measurements made directly from the speech spectrum on recordings of Korean and English collected from 45 speakers. They consist of two separate groups of native Korean and native English speakers, each including both male and female speakers. Based on the three acoustic parameters of open quotient (OQ), first-formant bandwidth (B1), and spectral tilt (ST), comparisons were made between the speech of males and females, between the speech of native Korean and native English speakers, and between Korean and English produced by native Korean speakers. Acoustic analysis of the experimental data indicates that some or all glottal parameters play a crucial role in differentiating the speech groups, despite substantial interspeaker variations. Statistical analysis of the Korean data indicates prosodic strengthening with respect to the acoustic parameters B1 and OQ, suggesting acoustic enhancement in terms of the degree of glottal abduction and the glottal closure during a vibratory cycle.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.5 no.4
• /
• pp.363-371
• /
• 2010

In an effort to enhance the quality of feature vector classification and thereby reduce the recognition error rate for the speaker-independent speech recognition, we study the effect of spectral tilt on the Fourier magnitude spectrum en route to the extraction of MFCC. The effect of FIR filtering on the speech signal on the speech recognition is also investigated in parallel. Evaluation of the proposed methods are performed by two independent ways of the Fisher discriminant objective function and speech recognition test by hidden Markov model with fuzzy vector quantization. From the experiments, the recognition error rate is found to show about 10% relative improvements over the conventional method by an appropriate choice of the tilt factor.

• Journal of Biomedical Engineering Research
• /
• v.30 no.2
• /
• pp.169-178
• /
• 2009

In this study, we suggest Korean nonlinear fitting formula (KNFF) to maximize speech intelligibility for digital hearing aids based on NAL-NL1 (NAL-nonlinear, version 1). KNFF was derived from the same procedure which is used for deriving NAL-NL1. KNFF consider the long-term average speech spectrum of Korean instead of English because the frequency characteristic of Korean is different from that of English. New insertion gains of KNFF were derived using the SII (speech intelligibility index) program provided by ANSI. In addition, the insertion gains were modified to maximize the intelligibility of high frequency words. To verify effect of the new fitting gain, we performed speech discrimination test (SDT) and preference test using the hearing loss simulator from NOISH. In the SDT, a word set as test material consists of 50 1-syllable word generally used in hearing clinic. As a result of the test, in case of moderate hearing loss with severe loss on high frequency, the SDT scores of KNFF was more improved about 3.2% than NAL-NLl and about 6% in case of the sever hearing loss. Finally we have obtained the result that it was the effective way to increase gain of mid-high frequency bands and to decrease gain of low frequency bands in order to maximize speech intelligibility of Korean.

• 전자공학회논문지 IE
• /
• v.49 no.2
• /
• pp.9-17
• /
• 2012

One of the problems on the application of the speech recognition system in the real world is the degradation of the performance by acoustical distortions. The most important source of acoustical distortion is the additive noise. This paper describes a spectral compensation technique based on a spectral peak enhancement scheme followed by an efficient noise subtraction scheme for noise robust speech recognition. The proposed methods emphasize the formant structure and compensate the spectral tilt of the speech spectrum while maintaining broad-bandwidth spectral components. The recognition experiments was conducted using noisy speech corrupted by white Gaussian noise, car noise, babble noise or subway noise. The new technique reduced the average error rate slightly under high SNR(Signal to Noise Ratio) environment, and significantly reduced the average error rate by 1/2 under low SNR(10 dB) environment when compared with the case of without spectral compensations.

• Phonetics and Speech Sciences
• /
• v.6 no.4
• /
• pp.91-99
• /
• 2014

Speaking styles tend to have an influence on spectral characteristics of produced speech. There are not many studies on the spectral characteristics of speech because of complicated processing of too much spectral data. The purpose of this study was to examine spectral characteristics and formant bandwidths of English vowels produced by nine American males with different speaking styles: clear or conversational styles; high- or low-pitched voices. Praat was used to collect pitch-corrected long-term averaged spectra and bandwidths of the first two formants of eleven vowels in the speaking styles. Results showed that the spectral characteristics of the vowels varied systematically according to the speaking styles. The clear speech showed higher spectral energy of the vowels than that of the conversational speech while the high-pitched voice did the same over the low-pitched voice. In addition, front and back vowel groups showed different spectral characteristics. Secondly, there was no statistically significant difference between B1 and B2 in the speaking styles. B1 was generally lower than B2 when reflecting the source spectrum and radiation effect. However, there was a statistically significant difference in B2 between the front and back vowel groups. The author concluded that spectral characteristics reflect speaking styles systematically while bandwidths measured at a few formant frequency points do not reveal style differences properly. Further studies would be desirable to examine how people would evaluate different sets of synthetic vowels with spectral characteristics or with bandwidths modified.

• The Journal of the Acoustical Society of Korea
• /
• v.29 no.1
• /
• pp.82-90
• /
• 2010

This paper proposes an efficient method to enhance the quality of synthesized speech in HMM based speech synthesis system. The proposed method trains spectral parameters and excitation signals using Gaussian mixture model, and estimates appropriate excitation signals from spectral parameters during the synthesis stage. Both WB-PESQ and MUSHRA results show that the proposed method provides better speech quality than conventional HMM based speech synthesis system.

• The Journal of the Acoustical Society of Korea
• /
• v.22 no.1E
• /
• pp.43-48
• /
• 2003

The effects of linear acoustic channels have been analyzed and compensated at mel-frequency feature domain. Unlike popular RASTA filtering our approach incorporates separate filters for each mel-frequency band, which results in better recognition performance for heavy-reverberated speeches.

• The Journal of the Acoustical Society of Korea
• /
• v.18 no.5
• /
• pp.12-20
• /
• 1999

In this paper, we compared speaker verification performance of the speech data collected in clean environment and in channel environment. For the improvement of the performance of speaker verification gathered in channel, we have studied on the efficient feature parameters in channel environment and on the preprocessing. Speech DB for experiment is consisted of Korean doublet of numbers, considering the text-prompted system. Speech features including LPCC(Linear Predictive Cepstral Coefficient), MFCC(Mel Frequency Cepstral Coefficient), PLP(Perceptually Linear Prediction), LSP(Line Spectrum Pair) are analyzed. Also, the preprocessing of filtering to remove channel noise is studied. To remove or compensate for the channel effect from the extracted features, cepstral weighting, CMS(Cepstral Mean Subtraction), RASTA(RelAtive SpecTrAl) are applied. Also by presenting the speech recognition performance on each features and the processing, we compared speech recognition performance and speaker verification performance. For the evaluation of the applied speech features and processing methods, HTK(HMM Tool Kit) 2.0 is used. Giving different threshold according to male or female speaker, we compare EER(Equal Error Rate) on the clean speech data and channel data. Our simulation results show that, removing low band and high band channel noise by applying band pass filter(150～3800Hz) in preprocessing procedure, and extracting MFCC from the filtered speech, the best speaker verification performance was achieved from the view point of EER measurement.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.16 no.5
• /
• pp.239-246
• /
• 2023

In this paper, we propose an open-loop algorithm to classify speech and music signals using the spectral flux parameters and Mel Frequency Cepstral Coefficients(MFCC) parameters for the audio coder. To increase responsiveness, the MFCC was used as a short-term feature parameter and spectral fluxes were used as a long-term feature parameters to improve accuracy. The overall voice/music signal classification decision is made by combining the short-term classification method and the long-term classification method. The Gaussian Mixed Model (GMM) was used for pattern recognition and the optimal GMM parameters were extracted using the Expectation Maximization (EM) algorithm. The proposed long-term and short-term combined speech/music signal classification method showed an average classification error rate of 1.5% on various audio sound sources, and improved the classification error rate by 0.9% compared to the short-term single classification method and 0.6% compared to the long-term single classification method. The proposed speech/music signal classification method was able to improve the classification error rate performance by 9.1% in percussion music signals with attacks and 5.8% in voice signals compared to the Unified Speech Audio Coding (USAC) audio classification method.