• Journal of Broadcast Engineering
• /
• v.17 no.3
• /
• pp.519-528
• /
• 2012

Voice activity detection (VAD) is generally conducted by extracting features from the acoustic signal and a decision rule. The performance of such VAD algorithms driven by the input acoustic signal highly depends on the acoustic noise. When video signals are available as well, the performance of VAD can be enhanced by using the visual information which is not affected by the acoustic noise. Previous visual VAD algorithms usually use single visual feature to detect the lip activity, such as active appearance models, optical flow or intensity variation. Based on the analysis of the weakness of each feature, we propose to combine intensity change measure and the optical flow in the mouth region, which can compensate for each other's weakness. In order to minimize the computational complexity, we develop simple measures that avoid statistical estimation or modeling. Specifically, the optical flow is the averaged motion vector of some grid regions and the intensity variation is detected by simple thresholding. To extract the mouth region, we propose a simple algorithm which first detects two eyes and uses the profile of intensity to detect the center of mouth. Experiments show that the proposed combination of two simple measures show higher detection rates for the given false positive rate than the methods that use a single feature.

• The Journal of the Acoustical Society of Korea
• /
• v.32 no.5
• /
• pp.369-376
• /
• 2013

Using Lowson's acoustic analogy, low frequency noise of a wind turbine (WT) is predicted in time domain and the noise sources contributing to the low frequency noise is analyzed. To compute averaged pressure distribution on blades of the WT as noise source, XFOIL is utilized. The blade source domain is divided into several segments along the span direction to compute force exerted on air surrounding the blade segments, which is used as input for noise prediction. The noise sources are decomposed into three terms of force fluctuation, acceleration and velocity terms and are analyzed to investigate each spectral contribution. Finally, predicted spectra are compared with measured low frequency noise spectrum of a wind turbine in operation. It is found that the force fluctuation component contributes strongly in low frequency range with increasing wind speed.

• Phonetics and Speech Sciences
• /
• v.15 no.1
• /
• pp.9-17
• /
• 2023

The purpose of this study is to determine how the Korean diphthong /jo/ shows phonetic variation in various linguistic environments. The pronunciation of /jo/ is discussed, focusing on the relationship between phonetic variation and the distribution range of vowels. The location in a word (monosyllable, word-initial, word-medial, word-final) and word class (content word, function word) were analyzed using the speech of 10 female speakers of the Seoul Corpus. As a result of determining the frequency of appearance of /jo/ in each environment, the pronunciation type and word class were affected by the location in a word. Frequent phonetic reduction was observed in the function word /jo/ in the acoustic analysis. The word class did not change the average phonetic values of /jo/, but changed the distribution of individual tokens. These results indicate that the linguistic environment affects the phonetic distribution of vowels.

• International Journal of Naval Architecture and Ocean Engineering
• /
• v.6 no.3
• /
• pp.562-577
• /
• 2014

A reliable steady/transient hydro-elastic analysis is developed for flexible (composite) marine propeller blade design which deforms according to its environmental load (ship speed, revolution speed, wake distribution, etc.) Hydro-elastic analysis based on CFD and FEM has been widely used in the engineering field because of its accurate results however it takes large computation time to apply early propeller design stage. Therefore the analysis based on a boundary element method-Finite Element Method (BEM-FEM) Fluid-Structure Interaction (FSI) is introduced for computational efficiency and accuracy. The steady FSI analysis, and its application to reverse engineering, is designed for use regarding optimum geometry and ply stack design. A time domain two-way coupled transient FSI analysis is developed by considering the hydrodynamic damping ffects of added mass due to fluid around the propeller blade. The analysis makes possible to evaluate blade strength and also enable to do risk assessment by estimating the change in performance and the deformation depending on blade position in the ship's wake. To validate this hydro-elastic analysis methodology, published model test results of P5479 and P5475 are applied to verify the steady and the transient FSI analysis, respectively. As the results, the proposed steady and unsteady analysis methodology gives sufficient accuracy to apply flexible marine propeller design.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.29 no.5
• /
• pp.512-519
• /
• 2023

Noise pollution poses significant challenges to human well-being and marine ecosystems. It is primarily caused by the flow around ships and marine installations, emphasizing the need for accurate noise evaluation of flow noise to ensure environmental safety. Existing flow noise analysis methods for underwater environments typically use a hybrid method combining computational fluid dynamics and Ffowcs Williams-Hawkings acoustic analogy. However, this approach has limitations, neglecting near-field effects such as reflection, scattering, and diffraction of sound waves. In this study, an alternative using direct method flow noise analysis via the lattice Boltzmann method (LBM) is incorporated. The LBM provides a more accurate representation of the underwater structural boundaries and acoustic wave effects. Despite challenges in underwater environments due to numerical instabilities, a novel DM-TS LBM collision operator has been developed for stable implementations for hydroacoustic applications. This expands the LBM's applicability to underwater structures. Validation through flow noise analysis in pipe orifice demonstrates the feasibility of near-field analysis, with experimental comparisons confirming the method's reliability in identifying main pressure peaks from flow noise. This supports the viability of near-field flow noise analysis using the LBM.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.26 no.3
• /
• pp.297-307
• /
• 2020

In the study, energy flow analysis is performed to predict the performance of silencers. To date, deterministic approaches such as finite element method have been widely used for silencer analysis. However, they have limitations in analyzing large structures and mid-high frequency ranges due to unreasonable computational costs and errors. However, silencers used for ships and off-shore plants are much larger than those used in other engineering fields. Hence, energy governing equation, which is significantly efficient for systems with high modal density, is solved for silencers in ships and off-shore plants. The silencer is divided into two different acoustic media, air and absorption materials. The discontinuity of energy density at interfaces is solved via hypersingular integrals for the 3-D modified Helmholtz equation to analyze multi-domain problems with the energy flow boundary element method. The method is verified by comparing the measurements and analysis results for ship silencers over mid-high frequency ranges. The comparisons confirm good agreement between the measurement and analysis results. We confirm that the applied analysis method is useful for large silencers in mid-high frequency ranges. With the proven procedures, energy flow analysis can be performed for various types of silencer used in ships and off-shore plants in the first stage of the design.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.33 no.7
• /
• pp.18-25
• /
• 2005

In this study, the main focus is the simulation of acoustic wave scattering in trailing edge and the analysis of the generation mechanism of instability wave by the interaction of trailing edge, shear flow and initial disturbance. The numerical algorithm is based on CFD/CAA hybrid method with high-order computational aeroacoustic method. It is found that steady mean flow gradient terms play a crucial role on the generation of instability wave through the comparison of simulations of Simple Linearized Euler Equation and Full Linearized Euler Equation. Through the comparison with the results of Full Navier-Stokes Equation, it is reasonable and efficient to use the Full Linearized Euler Equation in the initial generation mechanism of the instability wave near the trailing edge.

• Journal of the Society of Naval Architects of Korea
• /
• v.44 no.2 s.152
• /
• pp.148-153
• /
• 2007

Two typical impact loadings, shock wave and gas bubble pulse, due to UNDEX(UNDerwater EXplosion), should be considered together for the closest response analysis of structure subjected to UNDEX to a reality. Since these two impact loadings have different response time bands, however, their response characteristics of structure are different from each other. It is impossible to consider these effectively under the current computational environment and the mathematical model has not yet been developed. Whereas Hicks model approximates the fluid-structure interaction due to gas bubble pulse as virtual mass effect, treating the flow by the response of gas bubble after shock wave as incompressible ideal fluid contrary to the compressible flow due to shock wave, Geers-Hunter model could make the closest response analysis of structure under UNDEX to a real one as a mathematical model considering the fluid-structure interaction due to shock wave and gas bubble pulse together using acoustic wave theory and DAA(Doubly Asymptotic Approximation). In this study, the application and effectiveness of integrated dynamic response analysis of submerged structure was examined with the analysis of the shock wave and gas bubble pulse together.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.41 no.6
• /
• pp.261-269
• /
• 2004

In this paper, an efficient algorithm for computing in the MPEG/audio Layer Ⅲ (MP3) encoder is proposed. The proposed algerian performs a full-band 1024-point FFT by computing 32-point FFT's of 32 subband outputs. To reduce the aliasing caused by the analysis filter bank, an aliasing cancellation butterfly is developed. A major benefit of the proposed algorithm is the computational saving. By using the proposed algorithm, it is possible to save 40~50% of computations for FFT, which results in about 20% reduction of the PAM-2 complexity.

• New & Renewable Energy
• /
• v.3 no.3
• /
• pp.45-53
• /
• 2007

This paper demonstrates a computational method in predicting aerodynamic noise generated from wind turbines. Low frequency noise due to displacement of fluid and leading fluctuation, according to the blade passing motion, is modelled on monopole and dipole sources. They are predicted by Farassat 1A equation. Airfoil self noise and turbulence ingestion noise are modelled upon quadrupole sources and are predicted by semi-empirical formulas composed on the groundwork of Brooks et al. and Lowson. Aerodynamic flow in the vicinity of the blade should be obtained first, while noise source modelling need them as numerical inputs. Vortex Lattice Method(VLM) is used to compute aerodynamic conditions near blade. In the use of program X-foil [M.Drela] boundary layer characteristics are calculated to obtain airfoil self noise. Wind turbine blades are divided into spanwise unit panels, and each panel is considered as an independent source. Retarded time is considered, not only in low frequency noise but also In turbulence ingestion noise and airfoil self noise prediction. Numerical modelling is validated with measurement from NREL [AOC15/50 Turbine) and ETSU [Markham's VS45] wind turbine noise measurements.