• Journal of the Institute of Convergence Signal Processing
• /
• v.15 no.3
• /
• pp.91-96
• /
• 2014

For Sound source direction and separation method, Frequency Domain Binaural Model(FDBM) shows low computational cost and high performance for sound source separation. This method performs sound source orientation and separation by obtaining the Interaural Phase Difference(IPD) and Interaural Level Difference(ILD) in frequency domain. But the problem of reflection occurs in practical environment. To reduce this reflection, a method to simulate the sound localization of a direct sound, to detect the initial arriving sound, to check the direction of the sound, and to separate the sound is presented. Simulation results show that the direction is estimated to lie close within 10% from the sound source and, in the presence of the reflection, the level of the separation of the sound source is improved by higher Coherence and PESQ(Perceptual Evaluation of Speech Quality) and by lower directional damping than those of the existing FDBM. In case of no reflection, the degree of separation was low.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.12
• /
• pp.191-196
• /
• 2013

In this paper, the interaural intensity difference (IID)-based sounr source separation method in closely spaced stereo omnidirectional microphones is proposed. First, in order to improve the channel separability, a minimum variance distortionless response (MVDR) beamformer is employed to increase the intensity difference between stereo channels. After that, IID-based sound source separation method is applied. In order to evaluate the performance of the proposed method, source-to-distortion ratio (SDR), source-to-interference ratio (SIR), and sources-to-artifacts ratio (SAR), which are defined as objective evaluation criteria in stereo audio source separation evaluation campaign (SASSEC), are measured. As a result, it was shown from the objective evaluation that the proposed method outperforms a sound source separation method without applying a beamformer.

• Journal of the Korean Society for Precision Engineering
• /
• v.20 no.5
• /
• pp.116-123
• /
• 2003

These days, the researches related with the emotional robots are actively investigated and in progress. And human language, expression, action etc. are merged in the emotional robot to understand the human emotion. However, there are so many sound sources and background noise around the robot, that the robots should be able to separate the mixture of these sound sources into the original sound sources, moreover to understand the meaning of voice of a specific person. Also they should be able to turn or move to the direction of a specific person to observe his expression or action effectively. Until now, the researches on the localization and separation of sound sources have been so theoretical and computative that real-time processing is hardly possible. In this reason for the practical emotional robot, fast computation should be realized by using simple principle. In this paper the methods for detecting the direction of sound sources by using the phase difference between peaks on spectrums, and the separating the sound sources by using fundamental frequency and its overtones of human voice, are proposed. Also by using these methods, it is shown that the effective and real-time localization and separation of sound sources in living room are possible.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.4 no.2
• /
• pp.77-82
• /
• 2011

In this paper, we investigate source separation algorithms for stereo audio mixed using amplitude panning method. This source separation algorithms can be used in various applications such as up-mixing, speech enhancement, and high quality sound source separation. The methods in this paper estimate the panning angles of individual signals using the principal component analysis being applied in time-frequency tiles of the input signal and independently extract each signal through directional filtering. Performances of the methods were evaluated through computer simulations.

• The Journal of the Acoustical Society of Korea
• /
• v.21 no.4E
• /
• pp.146-155
• /
• 2002

We present a new technique for achieving source separation when given only a single charmel recording. The main idea is based on exploiting the inherent time structure of sound sources by learning a priori sets of time-domain basis functions that encode the sources in a statistically efficient manner. We derive a learning algorithm using a maximum likelihood approach given the observed single charmel data and sets of basis functions. For each time point we infer the source parameters and their contribution factors. This inference is possible due to the prior knowledge of the basis functions and the associated coefficient densities. A flexible model for density estimation allows accurate modeling of the observation, and our experimental results exhibit a high level of separation performance for simulated mixtures as well as real environment recordings employing mixtures of two different sources. We show separation results of two music signals as well as the separation of two voice signals.

• The Journal of the Acoustical Society of Korea
• /
• v.21 no.4
• /
• pp.146-146
• /
• 2002

We present a new technique for achieving source separation when given only a single channel recording. The main idea is based on exploiting the inherent time structure of sound sources by learning a priori sets of time-domain basis functions that encode the sources in a statistically efficient manner. We derive a learning algorithm using a maximum likelihood approach given the observed single channel data and sets of basis functions. For each time point we infer the source parameters and their contribution factors. This inference is possible due to the prior knowledge of the basis functions and the associated coefficient densities. A flexible model for density estimation allows accurate modeling of the observation, and our experimental results exhibit a high level of separation performance for simulated mixtures as well as real environment recordings employing mixtures of two different sources. We show separation results of two music signals as well as the separation of two voice signals.

• The Journal of the Acoustical Society of Korea
• /
• v.38 no.4
• /
• pp.427-432
• /
• 2019

In this paper, we propose a home monitoring system using sound event detection based on a bidirectional gated recurrent neural network for the hard-of-hearing. First, in the proposed system, packet loss concealment is used to recover a lost signal captured through wireless sensor networks, and reliable channels are selected using multi-channel cross correlation coefficient for effective sound event detection. The detected sound event is converted into the text and haptic signal through a harmonic/percussive sound source separation method to be provided to hearing impaired people. Experimental results show that the performance of the proposed sound event detection method is superior to the conventional methods and the sound can be expressed into detailed haptic signal using the source separation.

• Journal of Sensor Science and Technology
• /
• v.13 no.3
• /
• pp.224-229
• /
• 2004

In this paper, we propose a new Active Noise Control (ANC) system using a teacher forced Blind Source Separation (BSS) algorithm. The Blind Source Separation based on the Independent Component Analysis (ICA) separates the desired sound signal from the unwanted noise signal. In the proposed system, the BSS algorithm is used as a preprocessor of ANC system. Also, we develop a teacher forced BSS learning algorithm to enhance the performance of BSS. The teacher signal is obtained from the output signal of the ANC system. Computer experimental results show that the proposed ANC system in conjunction with the BSS algorithm effectively cancels only the ship engine noise signal from the linear and convolved mixtures with human voice.

• The Journal of the Acoustical Society of Korea
• /
• v.36 no.2
• /
• pp.130-135
• /
• 2017

In this paper, we propose an interface for interactive sound experience in the virtual concert hall. The proposed interface consists of two systems, called 'virtual acoustic position' and 'virtual active listening'. To provide these systems, we applied an artificial reverberation algorithm, multi-channel source separation and head-related transfer function. The proposed interface was implemented by using Unity. The interface provides the virtual concert hall to user through Oculus Rift, one of the virtual reality headsets. Moreover, we used Leap Motion as a control device to allow a user experience the system with free-hand. And user can experience the sound of the system through headphones.

• Journal of Broadcast Engineering
• /
• v.21 no.2
• /
• pp.169-179
• /
• 2016

Recently, MPEG-H has been standardizing for a multimedia coder in UHDTV (Ultra-High-Definition TV). Thus, the demand for not only channel-based audio contents but also object-based audio contents is more increasing, which results in developing a new technique of converting channel-based audio contents to object-based ones. In this paper, a non-uniform linear microphone array based source separation method is proposed for realizing such conversion. The proposed method first analyzes the arrival time differences of input audio sources to each of the microphones, and the spectral magnitudes of each sound source are estimated at the horizontal directions based on the analyzed time differences. In order to demonstrate the effectiveness of the proposed method, objective performance measures of the proposed method are compared with those of conventional methods such as an MVDR (Minimum Variance Distortionless Response) beamformer and an ICA (Independent Component Analysis) method. As a result, it is shown that the proposed separation method has better separation performance than the conventional separation methods.