• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.7
• /
• pp.2785-2799
• /
• 2020

We propose a novel speaker adaptation method using acoustic model clustering. The similarity of different speakers is defined by the cosine distance between their i-vectors (intermediate vectors), and various efficient clustering algorithms are applied to obtain a number of speaker subsets with different characteristics. The speaker-independent model is then retrained with the training data of the individual speaker subsets grouped by the clustering results, and an unknown speech is recognized by the retrained model of the closest cluster. The proposed method is applied to a large-scale speech recognition system implemented by a hybrid hidden Markov model and deep neural network framework. An experiment was conducted to evaluate the word error rates using Resource Management database. When the proposed speaker adaptation method using i-vector based clustering was applied, the performance, as compared to that of the conventional speaker-independent speech recognition model, was improved relatively by as much as 12.2% for the conventional fully neural network, and by as much as 10.5% for the bidirectional long short-term memory.

• Journal of the Korean Society of Laryngology, Phoniatrics and Logopedics
• /
• v.16 no.2
• /
• pp.135-139
• /
• 2005

Summary: This study evaluated the relationship between voice complaint and deviant vocal fold status with special regard to presbylarynx, in patients aged more than 60 years with pharyngeal-laryngeal complaint. The material consisted of clinical histories and images obtained by laryngoscopies of 75 patients aged more than 60 years, who had sought otorhinolaryngologic treatment. Indicative glottic characteristics of the presbylarynx, such as vocal fold bowing(VFB), prominence of vocal processes (PVP), and membranous spindle shaped glottic chink(MSC) and the presence or absence of voice complaint were analyzed Also, acoustic parameters such as fundamental frequency(Fo), jitter percent and shimmer percent were analyzed. VFB showed a strong correlation with voice complaint in male. Jitter and shimmer were correlated with VFB, PVP, MSC in female.

• Applied Chemistry for Engineering
• /
• v.17 no.1
• /
• pp.62-66
• /
• 2006

In this work, the properties of environmentally friendly functional panel made from waste aluminum were investigated. Product quality enhancement was pursued through an improved viscosity process, a mixing process by agitating, a foaming process, a cooling process, and a color addition process. An acoustic transmission attenuation test, a sound adsorption rate measurement test, and a foaming condition and scrap mixing test were implemented. As a result, the functional panel made from waste aluminum was ultra lightweight and had excellent properties such as soundproof, sound interception, and shielding harmful electromagnetic waves. Also, the functional panel showed low thermal conductivity (about 2.2 kcal/mh) and excellent heat-insulating property.

• 정보저장시스템학회:학술대회논문집
• /
• 2005.10a
• /
• pp.197-202
• /
• 2005

Current Optical Disk drive is dominating high share gradually in information storage device market through development of great skill. These technologies can achieve not only improvement of transmission speed but also elevation of recording/reproduction technology. However, these present state cause not only the increase of vibration in structural viewpoint but also problem of serious noise while the ODD becomes high speed/high performance. Specially, air-born noise is being bigger than structural-born noise as disk rotation speed increases gradually. The object of this research is that reduce the air-born noise in optical disk drive by reducing a quantity of flow by using a miniaturized muffler. The micro muffler is a miniaturized muffler. The muffler is used widely by solution to reduce air-born noise which is generated by flow. According to frequency band of the noise source, it can be applied by muffler of various forms. In this research, we examined the noise characteristics of the micro muffler and applied it by noise reduction solution of the ODD. It could get an excellent noise reduction in high frequency band by using the micro muffler than the opened case through the decrease of an inner flow. But it could not get a noise reduction in low frequency band through the decrease of an inner flow.

• The Journal of the Korea Contents Association
• /
• v.11 no.5
• /
• pp.138-147
• /
• 2011

While most Korean speech databases are developed for adults' speech, not for children's speech, there are various children's speech databases based on other languages. Because there are wide differences between children's and adults' speech in acoustic and linguistic characteristics, the children's speech database needs to be developed. In this paper, to find the differences between them in Korean, we built speech recognizers using HMM and tested them according to gender, age, and the presence of VTLN(Vocal Tract Length Normalization). This paper shows the speech recognizer made by children's speech has a much higher recognition rate than that made by adults' speech and using VTLN helps to improve the recognition rate in Korean.

• Phonetics and Speech Sciences
• /
• v.7 no.2
• /
• pp.85-94
• /
• 2015

This study is to investigate the characteristic of tonal elements in word utterance of 30 pre-school children. For the analyses, 240 utterances of 4 syllable words were processed to extract acoustic values and then the data was transformed into tonal height in order to examine the contour. The results show that the mean pitch of a note is $C4{\frac{1}{2}}(271.17Hz)$ and high and low pitched notes are $C5{\frac{1}{2}}(452.57Hz)$ and $G{\sharp}3{\frac{1}{2}}(192.54Hz)$. The pitch patterns of the 4 syllables measured at the frication and aspiration portion are $E4{\frac{1}{2}}-F4-B3{\frac{1}{2}}-A3$ and F4-E4-B3-A3. The pitch patterns of consonant clusters are $B3{\frac{1}{2}}-D4-B3{\frac{1}{2}}-A3{\frac{1}{2}}$ and $A{\sharp}3{\frac{1}{2}}-C4-A3-D4{\frac{1}{2}}$. The analyses of tonal elements in this study provide evidentiary data on tonal height helpful for developing melodic contour.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2002.11a
• /
• pp.140-143
• /
• 2002

In this study, we prepared ZnO/glass and $ZnO/SiO_{2}/Si$ thin film by Facing Targets Sputtering (FTS) system for Film Bulk Acoustic Resonator (FBAR). When the ZnO thin film applied to piezoelectric thin film, it requires good c-axis preferred orientation. And c-axis orientation has a remarkable difference with preparation conditions. Therefore, c-axis orientation must be significantly evaluated according to changing deposition conditions. Moreover, in order to prepare ZnO thin film with good crystallographic properties and progressive of efficiency of product process, the ZnO thin film should have to prepared as low temperature as possible. In this work, we prepared ZnO thin films on slide glass and $SiO_{2}/Si$ substrate. And the crystallographic characteristics of ZnO thin films on sputtering conditions were investigated by alpha-step and X-ray diffraction.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2005.05b
• /
• pp.10-14
• /
• 2005

Zinc Oxide(ZnO) thin films on Si (100) substrates were deposited by RF magnetron reactive sputtering. The characteristics of zinc oxide thin films with changing sputtering conditions such as argon/oxygen gas ratios, RF power, and substrate temperature, chamber pressure and target-substrate distance were investigated. To analyze a crystallographic properties of the films, $\theta/2{\theta}$ mode X -ray diffraction, SEM, and AFM analyses. C-axis preferred orientation, resistivity, and surface roughness highly depended on Ar/$O_2$ gas ratios. The resistivity of ZnO thin films rapidly increased with increasing oxygen ratio and the resistivity value of $9{\times}10^7\;{\Omega}cm$ was obtained at a working pressure of 10 mTorr with Ar/$O_2$=50/50. The surface roughness was also improved with increasing oxygen ratio and the ZnO films deposited with Ar/$O_2$=50/50 showed the excellent roughness value of $28.7{\AA}$.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.17 no.1
• /
• pp.101-107
• /
• 2018

In this study, the analysis of radiation noise and chattering noise of vehicle seat recliner was conducted through testing and analysis. First, operating noise was measured by seat back frame and recliner, and chattering noise was confirmed. Next, the transient dynamic analysis was performed, and the result was mapped to the acoustic analysis. Finally, the test and analysis were compared and analyzed. The results are as follow. First, it was found that the peaks appeared in common in the range of 620~650Hz, 1,240~1,290 Hz, and 1,840~1,940 Hz. It was judged that the dynamic characteristics of the recliner system overlapped with the rotation component of the motor to cause amplification of noise and vibration. Next, as a result of imaging the radiation noise analysis, it was judged that the noise radiated in the forward and backward direction has a greater influence than the direction of the rotation axis when the ear position of the person is taken as a reference.

• Journal of the Korean Society of Propulsion Engineers
• /
• v.20 no.6
• /
• pp.61-69
• /
• 2016

This study is the results related to the combustion instability phenomenon with respect to combustor length and thermal power as variables in dual swirling combustor configuration. Especially, the beating phenomena having the insensitive resonance frequency of relatively constant peaks are observed when the combustor lengths increase in a lower power regime. This beating phenomenon might be occurred due to the interacting behaviors of pilot and main burners with different periods. Therefore, such insensitive response seems to be a result of the beating phenomenon with interaction between the pilot and main flames even though the combustor lengths are increased.