Music summarization refers to a technique which automatically extracts the most important and representative segments in music content. In this paper, we propose and evaluate a technique which provides the repeated part in music content as music summary. For extracting a repeated segment in music content, the proposed algorithm uses the weighted sum of similarity measures based on multi-level vector quantization for fixed-length summary or optimal-length summary. For similarity measures, count-based similarity measure and distance-based similarity measure are proposed. The number of the same codeword and the Mahalanobis distance of features which have same codeword at the same position in segments are used for count-based and distance-based similarity measure, respectively. Fixed-length music summary is evaluated by measuring the overlapping ratio between hand-made repeated parts and automatically generated ones. Optimal-length music summary is evaluated by calculating how much automatically generated music summary includes repeated parts of the music content. From experiments we observed that optimal-length summary could capture the repeated parts in music content more effectively in terms of summary length than fixed-length summary.

Donghae City - Ulleung Island Line (DC-UI Line) is a representative line for underwater and geoacoustic modeling in the middle western East Sea. In this line, an integrated model of P-wave velocity is proposed for a low-frequency range target (<200 Hz), based on high-resolution seismic profiles (2 - 7 kHz sonar and air-gun), shallow and deep cores (grab, piston, and Portable Remote Operated Drilling), and outcrop geology (Tertiary rocks and the basement on land). The basement comprises 3 geoacoustic layers of P-wave velocity ranging from 3750 to 5550 m/s. The overlying sediments consist of 7 layers of P-wave velocities ranging from 1500 to 1900 m/s. The bottom model shows that the structure is very irregular and the velocity is also variable with both vertical and lateral extension. In this area, seabed and underwater acousticians should consider that low-frequency acoustic modeling is very range-dependent and a detailed geoacoustic model is necessary for better modeling of acoustic propagation such as long-range surveillance of submarines and monitoring of currents.

SBN (Structure-Borne Noise) reduction in resiliently mounted machineries are predicted by using mass-spring model and wave model. In mass-spring model, mount is modeled as a spring, while in wave model, mount is considered as an equivalent elastic rod for taking account into longitudinal wave propagation. The predictions for SBN reduction through mounts are compared to the measurements for four different pumps. It is found that the mass-spring model is valid only in low frequency range below few hundred Hz, while for high frequency ranges longitudinal wave propagation in the mount must be considered to explain the measurements. It is also shown that impedance of the floor slightly affects low frequency behaviour in mass-spring and wave model below 50 Hz - 80 Hz, so that in engineering practice the effect of floor impedance may be neglected in computing mount performance.

This paper presents the automatic music emotion classification on acoustic data. A three-level structure is developed. The low-level extracts the timbre and rhythm features. The middle-level estimates the indication functions that represent the emotion probability of a single analysis unit. The high-level predicts the emotion result based on the indication function values. Experiments are carried out on 695 homogeneous music pieces labeled with four emotions, including pleasant, calm, sad, and excited. Three machine learning methods, GMM, MLP, and SVM, are compared on the high-level. The best result of 90.16% is obtained by MLP method.

Using the neural network model for oriented principal component analysis (OPCA), we propose a solution to the data least squares (DLS) problem, in which the error is assumed to lie in the data matrix only. In this paper, we applied this neural network model to channel equalization. Simulations show that the neural network based DLS outperforms ordinary least squares in channel equalization problems.