• The Journal of the Acoustical Society of Korea
• /
• v.39 no.5
• /
• pp.497-504
• /
• 2020

Constructing speaker embeddings in speaker verification is an important issue. In general, a self-attention mechanism has been applied for speaker embedding encoding. Previous studies focused on training the self-attention in a high-level layer, such as the last pooling layer. In this case, the effect of low-level layers is not well represented in the speaker embedding encoding. In this study, we propose Masked Cross Self-Attentive Encoding (MCSAE) using ResNet. It focuses on training the features of both high-level and low-level layers. Based on multi-layer aggregation, the output features of each residual layer are used for the MCSAE. In the MCSAE, the interdependence of each input features is trained by cross self-attention module. A random masking regularization module is also applied to prevent overfitting problem. The MCSAE enhances the weight of frames representing the speaker information. Then, the output features are concatenated and encoded in the speaker embedding. Therefore, a more informative speaker embedding is encoded by using the MCSAE. The experimental results showed an equal error rate of 2.63 % using the VoxCeleb1 evaluation dataset. It improved performance compared with the previous self-attentive encoding and state-of-the-art methods.

• The Journal of the Acoustical Society of Korea
• /
• v.39 no.5
• /
• pp.441-446
• /
• 2020

In this paper, we propose the α-Feature Map Scaling (α-FMS) method which extends the FMS method that was designed to enhance the discriminative power of feature maps of deep neural networks in Speaker Verification (SV) systems. The FMS derives a scale vector from a feature map and then adds or multiplies them to the features, or sequentially apply both operations. However, the FMS method not only uses an identical scale vector for both addition and multiplication, but also has a limitation that it can only add a value between zero and one in case of addition. In this study, to overcome these limitations, we propose α-FMS to add a trainable parameter α to the feature map element-wise, and then multiply a scale vector. We compare the performance of the two methods: the one where α is a scalar, and the other where it is a vector. Both α-FMS methods are applied after each residual block of the deep neural network. The proposed system using the α-FMS methods are trained using the RawNet2 and tested using the VoxCeleb1 evaluation set. The result demonstrates an equal error rate of 2.47 % and 2.31 % for the two α-FMS methods respectively.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.45 no.1
• /
• pp.35-51
• /
• 2017

Street gardens, a series of streetscape improvement projects led by Seoul City Government, are initiated for the purpose of providing aesthetic satisfaction and mental refreshment to pedestrians. In order to investigate whether street gardens indeed promote the psychological health of the users, questionnaire surveys were conducted on three selected street gardens - at Gangnam-daero, Digital-ro, and Teheranro - and their comparison sites located on the same streets, which have a similar physical environment but without a street garden. The survey questionnaires, based on Attention Restoration Theory, were composed of Perceived Restorativeness Scale-11 with the eleven individual questions grouped into four categories: 'Fascination', 'Being away', 'Coherence', and 'Scope'. The survey questionnaires also ask about physical components that promote psychological improvement in the aforementioned categories. The collected data was analyzed with factor analysis, reliability analysis, and independent t-test. The results suggested that street gardens had a relatively positive effect on the psychological restorativeness of the users. In particular, they gave fascination and interest to the users. However, they did not offer a feeling of being away to the users, which revealed the limitation in the psychological improvement effect of street gardens. The physical components of the street garden that have led the psychological restorativeness effect were wooden bench, tree, and flower. This result corresponds to an extant theory that natural factors have a positive effect on the psychological restorativeness within a hardscape. This research will shed light on the planning and design guidelines for the street garden project.

• Korean Journal of Childcare and Education
• /
• v.1 no.1
• /
• pp.125-146
• /
• 2005

This study examined the interactive effects of physical environment of childcare centers and teachers' creative teaching approaches on the level of young children's creativity. To do so, the study conducted an assessment called TCAM(Thinking Creativity in Action and Movement) that was developed by Torrance to 182 young children. Also, it utilized questionnaires to ask 28 teachers concerning physical environment of childcare centers and their creative teaching approaches level. The findings of this study were: First, the gender of young children didn't affect the level of young children's creativity but the age of young children positively affected the level of their imagination that is the subordinate area of creativity. Second, the high level of physical environment of childcare centers positively affected the level of young children's creativity. Third, teachers' actively creative teaching approaches positively affected the level of young children's creativity. Forth, both physical environment of childcare centers and teachers' creative teaching approaches interactively and positively influenced the level of young children's creativity. Fifth, both physical environment of childcare centers and teachers' creative teaching approaches positively affected the level of young children's creativity. The result of this study implies that the level of young children's creativity increases when childcare centers demonstrate high quality of physical environment and teachers deliver creative teaching approaches actively.

• Korea and Global Affairs
• /
• v.2 no.1
• /
• pp.93-112
• /
• 2018

The Pyeongchang 2018 Olympic Winter Games is a good example of functionalism in integration theories. President Moon Jae-in is extremely lucky to play host to the Winter Olympics. Moon should be particularly happy to have declared the 23rd Winter Games open, because a handful of North Korean athletes marched into the Pyeongchang Stadium as members of a joint team from "Corea," the result of his strenuous efforts to have the North participate in the world festival of sports on snow and ice. But the president of this divided nation hardly draws envy from other world leaders, as he is faced with the daunting task of accommodating the selfish positions of surrounding powers concerning North Korean nuclear and missile threats. North Korea, a trivial competitor in winter sports, scored big outside the games' sporting arenas by inviting President Moon to summit talks in Pyongyang. As a precondition for a 2018 summit, Pyongyang will first ask for the cessation of the annual joint Korea-US military exercises. President Moon invested a lot in the Olympic delegates from the North. Korea's leader will now have to start a truly difficult game which will require the best of best strategies as well as a great deal of wisdom and tenacity not only to deal with the weapons of mass destruction-toting North Koreans, but also with allies. On the other hand, Moon needs to make the effort to reset domestic politics with tolerance and compromise, so he can better concentrate on the conundrum of North Korean nuclear and missile threats.

• The Journal of the Acoustical Society of Korea
• /
• v.33 no.2
• /
• pp.87-93
• /
• 2014

If one identifies the detailed distribution of pressure and axial velocity at a source plane, the position and strength of major noise sources can be known, and the propagation characteristics in axial direction can be well understood to be used for the low noise design. Conventional techniques are usually limited in considering the constant source characteristics specified on the whole source surface; then, the source activity cannot be known in detail. In this work, a method to estimate the pressure and velocity field distribution on the source surface with high spatial resolution is studied. The matrix formulation including the evanescent modes is given, and the nearfield measurement method is proposed. Validation experiment is conducted on a wide duct system, at which a part of the source plane is excited by an acoustic driver in the absence of airflow. Increasing the number of evanescent modes, the prediction of pressure spectrum becomes further precise, and it has less than -25 dB error with 26 converged evanescent modes within the Helmholtz number range of interest. By using the converged modal amplitudes, the source parameter distribution is restored, and the position of the driver is clearly identified at kR = 1. By applying the regularization technique to the restored result, the unphysical minor peaks at the source plane can be effectively suppressed with the filtering of the over-estimated pure radial modes.

• The Journal of the Acoustical Society of Korea
• /
• v.36 no.3
• /
• pp.172-178
• /
• 2017

Ultrasound has been widely used for biological and medical applications including induction of cell death, but a precise mechanism of induced cell death by ultrasound is controversial. In this study, an irradiation system with 40 kHz ultrasound was developed for a suitable cell death test of a representative unicellular organism, yeast, and used to study the biological effect of ultrasound on inducing cell death. Potassium Iodide (KI) dosimetry was used to devise an optimal system that successfully delivers 40 kHz ultrasound and produces reactive oxygen species in a 1.5 ml Eppendorf tube. Cell death was observed in an ultrasound transmission time-dependent fashion in this system. Thermal effect during irradiation was not observable in ultrasound induced cell death. Co-treatment of 40 kHz ultrasound and hydrogen peroxide showed a synergistic effect in inducing cell death. This finding suggests that 40 kHz ultrasound is related to reactive oxygen species formation. However, NAC (N-acetyl-L-cysteine) oxygen scavenger slightly inhibited the cell death by 40 kHz ultrasound. It was also found that 40 kHz ultrasound induced cell death was slightly inhibited by inhibitors of necrosis or apoptosis (glycyrrhizin or zVAD-fmk). This study suggests that cell death induced by 40 kHz ultrasound may not be exclusively related to reactive oxygen species formation and thermal effects in irradiated yeast cells.

• The Journal of the Acoustical Society of Korea
• /
• v.38 no.6
• /
• pp.651-660
• /
• 2019

The present study aims to find out the necessary speech sound level which can satisfy with the speech intelligibility in a noisy classroom environments. For this, auralized materials were made to undertake listening tests with 27 people. Speech intelligibility tests were carried out using both Consonant-Vowel-Consonant (CVC) and Phonetically Balanced Words (PBW) methods. Signal to noise ratio was changed by 5 dB for each test. As a result, it was found that speech intelligibilities are increasing with larger Signal to Noise Ratio (SNR). It was also found that there is a lot of difference of speech intelligibilities by SNR for syllables (CVC) with the Reverberation Time (RT) of 1.5 s. However, any significant difference was not found for words (PBW) in the case with RTs of below 0.8 s. Also, it was revealed through the 2-way analysis of variance (ANOVA) test that SNR is the only attentive factor which can affect the Korean speech intelligibilities for both PBW and CVC methods. Therefore, RTs below 0.8 s could be the acoustic criteria for classroom which can minimize the effects of noise. In the case with RTs larger than 0.8 s, much larger SNR is needed to give sufficient speech intelligibility.

• The Journal of the Acoustical Society of Korea
• /
• v.34 no.1
• /
• pp.20-35
• /
• 2015

Underwater acoustic channel impulse responses (CIR) are influenced by sound speed profile (SSP), and the variation of CIR has significant effects on the performance of underwater acoustic communication systems. A significant change of SSP can occur within a short period, which must be considered during the design of underwater acoustic modems. This paper statistically analyzes the effect of the variation of SSP on the long-range acoustic signal propagation in shallow-water with thermocline using numerical modeling based on the data acquired from JACE13 experiment near Jeju island. The analysis result shows that CIR changes variously according to the SSP and the depth of the transmitter and receiver. We also found that when the transmitter and receiver are deeper, the variation of sound wave propagation pattern is smaller and signal level becomes higher. All CIR obtained in this study show that a series of bottom reflections due to downward refraction and small bottom loss in the shallow water with thermocline can be very important factor for long-range signal transmission and the performance of underwater acoustic communication system in time varying ocean environment can be very sensitive to the variation of SSP even for a short period of time.

• The Journal of the Acoustical Society of Korea
• /
• v.35 no.2
• /
• pp.92-101
• /
• 2016

There are many results in which acoustical conditions of a classroom play an important role for studying effects and academic achievement of students. However, there are very few guidelines or design proposals which could make appropriate acoustic environment when classrooms are built or renovated. The present study suggests various design proposals satisfying acoustic standards of classrooms based on theoretical calculation and acoustic field experiments. At first, minimum area of sound absorption was calculated which is required to satisfy the acoustic standard for domestic middle and high schools. Also, room acoustic measurements were carried out in order to investigate the acoustic performance of an existing classroom by changing interior finishing materials on ceiling and rear walls. As a result, it was revealed that reverberation time standard below 0.8 s can be acquired even if there is no sound absorption on ceiling which is a general practice executed in Korea. Specially, it was found that if partial area of ceiling would be treated as reflective with the ratio of sound absorption and reflection as 2:1, almost similar acoustic parameters of $C_{50}$, $D_{50}$, RASTI (Rapid Speech Transmission Index) and higher sound levels could be acquired in comparison with the case of entire sound absorption on ceiling.