• Proceedings of the IEEK Conference
• /
• 2000.07a
• /
• pp.253-256
• /
• 2000

This paper describes a DSP implementation of a real-time 3D sound localization algorithm with the use of a low power embedded DSP. A distinctive feature of this implementation is that the audible frequency band is divided into three, in accordance with the sound reflection and diffraction phenomena through different media from a certain sound source to human ears, and then in each subband a specific implementation procedure of the 3D sound localization is devised so as to operate real-time at a low frequency of 50MHz on a 16bit fixed-point DSP. Thus out DSP implementation can provide a listener with 3D sound effects through a headphone at low cost and low power consumption.

• Proceedings of the IEEK Conference
• /
• 2000.07a
• /
• pp.264-267
• /
• 2000

This paper describes a real-time 3D sound localization algorithm to be implemented with the use of a Bow power embedded DSP. This algorithm first divides the audible frequency band into three, on the basis of the analysis of the sound reflection and diffraction effects through different media from a certain sound source to human ears, and then in each subband a specific procedure is devised fur the 3D sound localization so as to operate real-time on a low power embedded DSP This algorithm aims at providing a listener with the 3D sound effects through a headphone at low cost and low power consumption.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.46 no.1
• /
• pp.71-78
• /
• 2009

Sound source localization systems in service robot applications estimate the direction of a human voice. Time delay information obtained from a few separate microphones is widely used for the estimation of the sound direction. Correlation is computed in order to calculate the time delay between two signals. In addition, PHAT weighting function can be applied to significantly improve the accuracy of the estimation. However, FFT and IFFT operations in the PHAT weighting function occupy more than half of the area of the sound source localization system. Thus efficient FFT and IFFT designs are essential for the IP implementation of sound source localization system. In this paper, we propose an efficient FFT/IFFT design method based on the characteristics of human voice.

• The Journal of the Acoustical Society of Korea
• /
• v.43 no.2
• /
• pp.178-183
• /
• 2024

This study introduces an enhanced sound source localization technique, bolstered by a data-driven deep learning approach, to improve the precision and accuracy of direction of arrival estimation. Focused on refining Time Difference Of Arrival (TDOA) based sound source localization, the research hinges on accurately estimating TDOA from cross-correlation functions. Accurately estimating the TDOA still remains a limitation in this research field because the measured value from actual microphones are mixed with a lot of noise. Additionally, the digitization process of acoustic signals introduces quantization errors, associated with the sampling frequency of the measurement system, that limit the precision of TDOA estimation. A deep learning-based approach is designed to overcome these limitations in TDOA accuracy and precision. To validate the method, we conduct comprehensive evaluations using both two and three-microphone array configurations. Moreover, the feasibility and real-world applicability of the suggested method are further substantiated through experiments conducted in an anechoic chamber.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.46 no.2
• /
• pp.114-119
• /
• 2009

Sound source localization systems in service robot applications estimate the direction of a human voice. Time delay information obtained from a few separate microphones is widely used for the estimation of the sound direction. Correlation is computed in order to calculate the time delay between two signals. Inverse cosine is used when the position of the maximum correlation value is converted to an angle. Because of nonlinear characteristic of inverse cosine, the accuracy of the computed angle is varied depending on the position of the specific sound source. In this paper, we propose an efficient sound source localization system using angle division. By the proposed approach, the region from $0^{\circ}$ to $180^{\circ}$ is divided into three regions and we consider only one of the three regions. Thus considerable amount of computation time is saved. Also, the accuracy of the computed angle is improved since the selected region corresponds to the linear part of the inverse cosine function. By simulations, it is shown that the error of the proposed algorithm is only 31% of that of the conventional a roach.

• The Journal of the Acoustical Society of Korea
• /
• v.27 no.4
• /
• pp.207-214
• /
• 2008

To create 3D sound, we usually use two methods which are two channels or multichannel sound systems. Because of cost and space problems, we prefer two channel sound system to multi-channel. Using a headphone or two speakers, the most typical method to create 3D sound effects is a technology of head related transfer function (HRTF) which contains the information that sound arrives from a sound source to the ears of the listener. But it causes a problem to localize a sound source around a certain places which is called cone-of-confusion. In this paper, we proposed the new algorithm to reduce the confusion of sound image localization. HRTF grouping and psychoacoustics theory are used to boost the spectral cue with spectrum difference among each directions. Informal listening tests show that the proposed method improves the front-back sound localization characteristics much better than conventional methods.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.8C
• /
• pp.1142-1148
• /
• 2004

A binaural filtering method using HRTF DB is generally used to make the headphone-based 3D sound. But it can make some confusion between front and back directions or between up and down directions due to the non-individual HRTF depending on each listener. To reduce the confusion of sound image localization, we propose a new method to boost the spectral cue by modifying HRTF spectra with spectrum difference between front and back directions. Informal listening tests show that the proposed method improves the front-back sound localization characteristics much better than the conventional methods

• The Journal of the Acoustical Society of Korea
• /
• v.36 no.6
• /
• pp.419-424
• /
• 2017

TDoA (Time Difference-of Arrival) or DoA (Direction-of-Arrival) can be used for source localization. However, the localizing performance is dependent on relative position between source and receivers, receivers' geometric structure, sound speed, and so on. In this paper we propose a source localization method with enhanced performance that combines multiple information. The proposed method uses the time TDoA, DoA and sound speed as variables. LM (Levenberg-Marquardt) method which is one of nonlinear optimizations is applied. The performances of the proposed method was evaluated by simulation. As result of simulation, the proposed method has the lower average localizing error performance than the previous method.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2005.11a
• /
• pp.200-204
• /
• 2005

The use of beamforming method and de-Dopplerization technique was applied in studying the rotating sound sources. Acoustic analysis of a moving sound source required that the measured sound signals be do-Dopplerized and restored as of the original emission signals. Two main issues of the signal reconstruction in time domain are addressed herein: First, to remove Doppler effect from the measured data and to restore the original emission data of the moving source. The difference of the time domain beamforming from the frequency domain beamforming was mentioned. Also, the time domain beamforming method is deployed in the test and the comparisons were made to the frequency domain results. The time domain signal reconstruction was numerically simulated prior to the application. To validate the de-Dopplerization Performance, the rotating Point sources were examined and localized by the use of a phased array of microphone. The application of prop-rotor was conducted in a hovering condition. The results of reconstructing time signals of rotating sources and its locations were shown in the power distribution maps. In the prop-rotor measurements, the acoustic source locations were successfully verified in varying positions for different frequencies of interest.

• The Journal of the Acoustical Society of Korea
• /
• v.20 no.3
• /
• pp.98-103
• /
• 2001

The binaural auditory system of human has the capability of differentiating the direction and distance of sound sources. This feature is well characterised in terms of the inter-aural intensity difference (IID), the inter-aural time difference (ITD) and/or the spectral shape difference (SSD) arising from the acoustic transfer of a sound source to the outer ears. This paper proposes an effective way of extracting the three sound perception factors (IID, ITD, SSD) from the head-related transfer functions (HRTF's) that depends on the direction and distance of the acoustic source from the listener. It includes the estimation method of the equivalent ITD and 1/3-octave band-based IID factors and their usage to locate a sound source in space. Subjective and objective tests were carried out to examine the effectiveness of the proposed methodology and its applicability to real sound systems. Those experimental results are illustrated in this paper.