• The Journal of the Acoustical Society of Korea
• /
• v.27 no.1E
• /
• pp.17-24
• /
• 2008

Blind speech separation (BSS) is well-known as a powerful technique for speech enhancement in many real world environments. In this paper, we propose a new application of BSS - acoustic echo cancellation (AEC) in a car environment. For this purpose, we develop a block-online BSS algorithm which provides robust separation than a batch version in changing environments with moving speakers. Simulation results using real world recordings show that the block-online BSS algorithm is very robust to speaker movement. When combined with AEC, simulation results using real audio recording in a car confirm the expectation that BSS improves double talk detection and echo suppression.

• Proceedings of the Korean Society of Tribologists and Lubrication Engineers Conference
• /
• 1993.12a
• /
• pp.77-84
• /
• 1993

Systems with journal bearings generally operate in large scale and under severe loading conditions such as steam generator turbines and internal combustion engines, in contrast to the machineries using rolling element bearings. Failure of the bearings in these machineries can result in the system breakdown. To avoid the time consuming repair and considerable economic loss, the detection of incipient failure in journal bearings becomes very important. In this experimental approach, acoustic emission monitoring is employed to the detection of incipient failure caused by intervention of foreign particles most probable in the journal bearing systems. It has been known that the intervention of foreign materials, insufficient lubrication and misassembly etc. are principal factors to cause bearing failure and distress. The experiment was conducted under such designed conditions as inserting alumina particles to the lubrication layer in the simulated journal bearing system. The results showed that acoustic emission could be an effective tool to detect the incipient failure in journal bearings.

• Proceedings of the KSME Conference
• /
• 2001.06a
• /
• pp.65-71
• /
• 2001

In this paper, acoustic omission technique(AE) has been applied to detect leak for heat exchanger by analyzing the characteristics of signal obtained from leak. It was confirmed that the characteristics of the signal generated by the turbulence of gas in the heat exchanger is narrow band signal having between 130-250KHz. Generally, the amplitude of leak signal is increased as the leak size increasing, but showed no significant change at frequency characteristic. Leak source location can be found by searching for the point of highest signal amplitude by comparing wi th several fired sensors.

• Proceedings of the KIEE Conference
• /
• 2002.07c
• /
• pp.1691-1693
• /
• 2002

Acoustic partial discharge detection has been used on GIS testing for more than 10 years. Sensitivities show that most of critical defect types can be detected with a reasonable signal-to-noise ratio. The paper reports on acoustic partial discharge detection comparable to sensitive electrical PD(partial discharge) measuring method.

• Proceedings of the KIEE Conference
• /
• 1988.07a
• /
• pp.967-970
• /
• 1988

This paper deals with automatic system of acoustic noise detection for realization of factory automation. The existing inspection process of failure products has mostly been executed in hand by rich-experienced workers. It is difficult to accomplish effectively or systematically the failure test of products owing to the diversality of ill-conditions. But the problem about it must be solved in viewpoint of cost down and factory automation in addition to the reliability of products. The necessity of automatic inspection system to inspect automatically undesirable acoustic noise of products which is one of the kinds of failure is suggested, and the procedure to develope it and the function of each system components are explained briefly.

• Journal of the Korean Society for Precision Engineering
• /
• v.16 no.3 s.96
• /
• pp.18-24
• /
• 1999

ThisPaper investigates the feasibility of using acoustic emission signal analysis for the detection of built-up edge during machining. Experiments were conducted on a CNC-lathe using conventional carbide insert tools under various cutting conditions. The cutting forces were also measured for comparisons. Experimental evidence is presented which indicates that the presence of a built-up edge can significantly affect the generation of acoustic emission in metal cutting. It is shown that under conditions in which a built-up edge is generated, the variation of $AE_{rms}$ signal with cutting speed can be quite different from the generally accepted linear, monotonic increase as previously reported. The feasibility of utilizing $AE_{rms}$ in built-up edge sensing is suggested.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.18 no.12
• /
• pp.1243-1249
• /
• 2008

Gear system has been widely used in industrial applications and unexpected failures of gears are not only extremely damaging but also leading to economic losses. So, early detection of fault is important for diagnosis machine condition. And acoustic emission is an efficient non-destructive testing technique fur the diagnosis of machine health and is useful technique far early detection of fault because it can find low-amplitude and high-frequency signal on account of high sensibility. Therefore, in this paper, the AE signal was measured and preprocessed using envelope analysis for gearbox with misalignment between pinion and gear. And then the gear misalignment's vibration characteristic were analyzed.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.32 no.5C
• /
• pp.478-486
• /
• 2007

Double-talk detection errors can result in either large residual echo or distorting the near-end talker's input speech. Thus accurate double-talk detection is an important problem in the acoustic echo canceller to improve the speech quality. In the double-talk detection algorithm using a cross-correlation coefficient, double-talk detection errors can occur in the initial convergence period of an adaptive filter or in noisy environment since the cross-correlation coefficient becomes large in such situations. In this paper, we propose a new double-talk detection algorithm based on the cross-correlation method using a normalized error signal power to reduce the double-talk detection errors. The experimental results have shown the performance improvement of an acoustic echo canceller as well as the noise-robustness of the proposed double-talk detector.

• The Journal of the Acoustical Society of Korea
• /
• v.42 no.5
• /
• pp.395-401
• /
• 2023

Sound Event Detection (SED) aims to identify not only sound category but also time interval for target sounds in an audio waveform. It is a critical technique in field of acoustic surveillance system and monitoring system. Recently, various models have introduced through Detection and Classification of Acoustic Scenes and Events (DCASE) Task 4. This paper explored how to design optimal parameters of DenseNet based model, which has led to outstanding performance in other recognition system. In experiment, DenseRNN as an SED model consists of DensNet-BC and bi-directional Gated Recurrent Units (GRU). This model is trained with Mean teacher model. With an event-based f-score, evaluation is performed depending on parameters, related to model architecture as well as model training, under the assessment protocol of DCASE task4. Experimental result shows that the performance goes up and has been saturated to near the best. Also, DenseRNN would be trained more effectively without dropout technique.

• Journal of Sensor Science and Technology
• /
• v.28 no.5
• /
• pp.318-322
• /
• 2019

In shooting training, an impact point identification system that uses the impact wave of the bullet to check the impact point in the target plate has been recently used. Acoustic sensors used in these systems must be able to detect shock waves of high sound pressure levels and be both waterproof and dustproof for rainy weather and dusty environments, respectively. In this study, membranes with excellent waterproof, dustproof, and sound transmitting characteristics were selected through a characteristics test; a protection cap was installed to install the selected materials. After coupling the produced protection cap to the acoustic sensor housing, the sensitivity and phase characteristics of the acoustic sensor were checked. Through the waterproof and dustproof test, the performances of its sensitivity and phase characteristics were confirmed. Finally, the normal shockwave of a 5.56 mm diameter bullet was measured using a shockwave detection signal collecting plate equipped with a prototype of the acoustic sensor at a 100 m firing range.