• Journal of Digital Convergence
• /
• v.14 no.7
• /
• pp.331-337
• /
• 2016

The purpose of this study is to provide a method to visualize the sound data to the three-dimensional image. The visualization of the sound data is performed according to the algorithm set after production of the text-based script that form the channel range of the sound data. The algorithm consists of a total of five levels, including setting sound channel range, setting picture frame for sound visualization, setting 3D image unit's property, extracting channel range of sound data and sound visualization, 3D visualization is performed with at least an operation signal input by the input device such as a mouse. With the sound files with the amount an animator can not finish in the normal way, 3D visualization method proposed in this study was highlighted that the low-cost, highly efficient way to produce creative artistic image by comparing the working time the animator with a study presented method and time for work. Future research will be the real-time visualization method of the sound data in a way that is going through a rendering process in the game engine.

• Smart Media Journal
• /
• v.11 no.1
• /
• pp.31-37
• /
• 2022

This paper proposes a study on applying signal processing and deep learning for sound recognition that detects sounds commonly heard in daily life (Screaming, Clapping, Crowd_clapping, Car_passing_by and Back_ground, etc.). In the proposed sound recognition, several techniques related to the spectrum of sound waves, augmentation of sound data, ensemble learning for various predictions, convolutional neural networks (CNN) deep learning, and two-dimensional (2-D) data are used for improving the recognition accuracy. The proposed sound recognition technology shows that it can accurately recognize various sounds through experiments.

• The Journal of the Acoustical Society of Korea
• /
• v.27 no.2E
• /
• pp.57-63
• /
• 2008

Stethoscopic auscultation is still one of the primary tools for the diagnosis of heart diseases due to its easy accessibility and relatively low cost. It is, however, a difficult skill to acquire. Many research efforts have been done on the automatic classification of heart sound signals to support clinicians in heart sound diagnosis. Recently, hidden Markov models (HMMs) have been used quite successfully in the automatic classification of the heart sound signal. However, in the classification using HMMs, there are so many heart sound signal types that it is not reasonable to assign a new class to each of them. In this paper, rather than constructing an HMM for each signal type, we propose to build an HMM for a set of acoustically-similar signal types. To define the classes, we use the KL (Kullback-Leibler) distance between different signal types to determine if they should belong to the same class. From the classification experiments on the heart sound data consisting of 25 different types of signals, the proposed method proved to be quite efficient in determining the optimal set of classes. Also we found that the class determination approach produced better results than the heuristic class assignment method.

• Proceedings of the IEEK Conference
• /
• 2000.07b
• /
• pp.1045-1048
• /
• 2000

This paper presents the application of the wavelet transform analysis and the neural network method to the phonocardiogram (PCG) signal. Heart sound is a acoustic signal generated by cardiac valves, myocardium and blood flow and is a very complex and nonstationary signal composed of many source. Heart sound can be discriminated normal heart sound and heart murmur. Murmurs have broader frequency bandwidth than the normal ones and can occur at random position of cardiac cycle. In this paper, we classified the group of heart sound as normal heart sound(NO), pre-systolic murmur(PS), early systolic murmur(ES), late systolic murmur(LS), early diastolic murmur(ED). And we used the wavelet transform to shorten artifacts and strengthen the low level signal. The ANN system was trained and tested with the back- propagation algorithm from a large data set of examples-normal and abnormal signals classified by expert. The best ANN configuration occurred with 15 hidden layer neurons. We can get the accuracy of 85.6% by using the proposed algorithm.

• Journal of the Korea Convergence Society
• /
• v.5 no.1
• /
• pp.35-41
• /
• 2014

An automatic speech recognition system is one of the popular research problems. There are many research groups working in this field for different language including Japanese. Japanese vowel recognition is one of important parts in the Japanese speech recognition system. The vowel classification system with the Mamdani fuzzy inference system was developed in this research. We tested our system on the blind test data set collected from one male native Japanese speaker and four male non-native Japanese speakers. All subjects in the blind test data set were not the same subjects in the training data set. We found out that the classification rate from the training data set is 95.0 %. In the speaker-independent experiments, the classification rate from the native speaker is around 70.0 %, whereas that from the non-native speakers is around 80.5 %.

• The Journal of the Acoustical Society of Korea
• /
• v.28 no.3
• /
• pp.260-267
• /
• 2009

It is very difficult to form a desired complex sound field at a designated region precisely as an application of acoustic arrays, which is one of important objects of array systems. To solve the problem, a filter design method was suggested, which employed the concept of an inverse method using the acoustical holography based on the boundary element method. In the acoustical holography used for the source identification, the measured field data are employed to reconstruct the vibro-acoustic parameters on the source surface. In the analogous problem of source array design, the desired field data at some specific points in the sound field was set as constraints and the volume velocity at the surface points of the source plane became the source signal to satisfy the desired sound field. In the filter design, the constraints for the desired sound field are set, first. The array source and given space are modelled by the boundary elements. Then, the desired source parameters are inversely calculated in a way similar to the holographic source identification method. As a test example, a target field comprised of a quiet region and a plane wave propagation region was simultaneously realized by using the array with 16 loudspeakers.

• International Journal of Automotive Technology
• /
• v.6 no.4
• /
• pp.367-374
• /
• 2005

This paper proposes a sound quality index to evaluate the vehicle interior noise. The index was developed using a correlation analysis of an objective measurement and a subjective evaluation data. First, the objective set of measurements was obtained at two specified driving conditions. One is from a wide-open test condition and the other is from a constant-speed test condition. At the same time, subjective evaluation was carried out using a score of ten scale where 17 test engineers participated in the experiment. The correlation analysis between the psycho-acoustic parameters derived from the objective measurement and the subjective evaluation was performed. The most critical factors at both test conditions were determined, and the corresponding equations for the sound quality were obtained from the multiple factor regression method. Finally, a comparative work between previous index and present index was performed to validate the effectiveness of the proposed index.

• Proceedings of the Acoustical Society of Korea Conference
• /
• spring
• /
• pp.149-152
• /
• 1999

This paper addresses major issues in localising sound sources in space using the experimental data set of head-related responses in the time or frequency domain. They come from the technical realisation steps for implementing the convolution of HRIR's with sound sources, the cross-talk cancellation for transaural filtering, the matched time delay compensation, etc. in real, those technical matters seem to be minor because they can be realised in off-line signal processing schemes. This paper puts much emphasis on what we misunderstood about the sets of HRTF's or HRIR's, More specifcaily, the sets of HRTF's or HRIR's of course supply relevant information to sound localisation but include much useless 'rubbish' that have made for us to fail to put spatial image into real souno signals such as voices and music's. This paper proposes possible reasons for such failure and, furthermore, introduces detained subjects that should be challenged so as to resolve them.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.26 no.3
• /
• pp.331-339
• /
• 2016

To improve the acoustic performance of sound absorbing materials, the thickness of the material should be increased or the sound absorbing material having an irregular surface shape should be used. In this study, the acoustic characteristics and methods to improve the acoustic performance of a sound absorbing system equipped with double layered polyester sound absorbing materials were investigated. The numerical model was set up and the results obtained from the model were compared with the actual measurement data. And, strategies to improve the acoustic performance of sound absorbing systems with double layered sound absorbing materials made of polyester with different configuration were shown. So, this study is expected to be usefully used at sites that require high acoustic absorption performance with minimal installation thickness to reduce sounds reflection in narrow spaces such as interior of subway tunnels or in noise barriers installed adjacent to rails.

• Journal of The Korean Association of Information Education
• /
• v.25 no.5
• /
• pp.705-712
• /
• 2021

This study aims to develop artificial intelligence education programs that can be easily applied in elementary schools according to the trend of the times called artificial intelligence. The training program designed the purpose and direction based on the analysis results of the needs of 70 elementary school teachers according to the steps of the ADDIE model. According to the survey, elementary school students developed a machine learning education program to set sound data as the theme of the most accessible in their daily lives and to learn the principles of artificial intelligence in solving problems using sound data in real life. These days, when the need for artificial intelligence education emerges, elementary machine learning education programs that solve daily life problems based on sound data developed in this study will lay the foundation for elementary artificial intelligence education.