A self-generator method for initial filters of SIMO-ICA applied to blind separation of binaural sound mixtures
ABSTRACT Blind separation of binaural mixed sounds using single-input multiple-output (SIMO)-model-based independent component analysis (SIMO-ICA) with self-generator for initial filter (SIMO-ICA-SG) is now being studied by the authors. This method contains frequency-domain ICA (FDICA-PB), single-talk detection, direction of arrival (DOA) estimation, head related transfer function (HRTF) matrix bank, and SIMO-ICA. This paper describes robustness of SIMO-ICA-SG against the mismatch of HRTF matrix bank. To evaluate it, the sound decomposition experiments are carried out under the real acoustic conditions. The experimental results reveal that the decomposition performance of the proposed method with mismatched HRTF matrix bank is superior to those of the conventional methods, and almost the same as those of the proposed method with matched one
SourceAvailable from: Kiyohiro Shikano[Show abstract] [Hide abstract]
ABSTRACT: We propose a new two-stage blind separation and deconvolution (BSD) algorithm for a convolutive mixture of speech, in which a new Single-Input Multiple-Output (SIMO)-model-based ICA (SIMO- ICA) and blind multichannel inverse filtering are combined. SIMO- ICA can separate the mixed signals, not into monaural source sig- nals but into SIMO-model-based signals from independent sources as they are at the microphones. After SIMO-ICA, a simple blind deconvolution technique for the SIMO model can be applied even when each source signal is temporally correlated. The simulation results reveal that the proposed method can successfully achieve the separation and deconvolution for a convolutive mixture of speech. Blind separation and deconvolution (BSD) of sources is an ap- proach taken to estimate original source signals using only the in- formation of mixed signals observed in each input channel. In the BSD framework, not only the source separation but also the decon- volution of the transmission channel characteristics are considered. For the BSD based on independent component analysis (ICA), var- ious methods have been proposed to deal with the separation and deconvolution for the convolutive mixture of independently, iden- tically distributed (i.i.d.) source signals (1, 2). These BSD meth- ods require the specific assumptions that the source signals are mu- tually independent and each source signal is also temporally inde- pendent. However, the latter assumption does not hold in many practical acoustic mixtures of sound signals such as speech. The application of the conventional ICA-based BSD to speech often yields the negative results, e.g., the separated speech is adversely decorrelated and whitened. In order to solve the problem, we have proposed a novel BSD approach that combines information- geometry theory and multichannel signal processing (3). In this approach, the BSD problem is resolved into two stages: new blind separation technique using a Single-Input Multiple-Output (SIMO)- model-based ICA (SIMO-ICA) and the deconvolution in the SIMO- model framework. In the previous report(3), we dealt with real-world data, but it is hard to say that we could make clear whether the proposed BSD can obtain exact source signals or not. With real-world data, it is difficult to evaluate the performance of the system accurately due to background noise, too long reverberation, and so on. In this paper, we give the objective indication of the performance in the first stage, and properly evaluate the performance of the pro- posed method using the artificial transmission channels. In ad- dition, we show that the proposed method can be regarded as a square FIR-type filter matrix, and we discuss the channel identifi- ability of such a system. The simulation results reveal that the pro- This work was partly supported by Core Research for Evolutional Sci- ence and Technology (CREST) in Japan. posed method can achieve the separation and deconvolution for a convolutive mixture of speech when we set the SIMO-ICA's filter length sufficiently long.
[Show abstract] [Hide abstract]
ABSTRACT: We propose a new Single-Input Multiple-Output (SIMO)-model-based ICA with information-geometric learning algorithm for high-fidelity blind source separation. The SIMO-ICA consists of multi-ple ICAs and a fidelity controller, and each ICA runs in parallel un-der the fidelity control of the entire separation system. The SIMO-ICA can separate the mixed signals, not into monaural source sig-nals but into SIMO-model-based signals from independent sources as they are at the microphones. Thus, the separated signals of SIMO-ICA can maintain the spatial qualities of each sound source. In order to evaluate its effectiveness, separation experiments are carried out under a reverberant condition. The experimental re-sults reveal that the signal separation performance of the proposed SIMO-ICA is the same as that of the conventional ICA-based method, and that The sound quality of the separated sound in SIMO-ICA is superior to that of the conventional method.
[Show abstract] [Hide abstract]
ABSTRACT: In this paper, single-input multiple-output (SIMO)-model-based blind source separation (BSS) is addressed, where unknown mixed source signals are detected at the microphones, and these signals can be separated, not into monaural source signals but into SIMO-model-based signals from independent sources as they are at the microphones. This technique is highly applicable to high-fidelity signal processing such as binaural signal processing. First, we provide an experimental comparison between two kinds of the SIMO-model-based BSS methods, namely, traditional frequency-domain ICA with projection-back processing (FDICA-PB), and SIMO-ICA recently proposed by the authors. Secondly, we propose a new combination technique of the FDICA-PB and SIMO-ICA, which can achieve a higher separation performance in comparison to two methods. The experimental results reveal that the accuracy of the separated SIMO signals in the simple SIMO-ICA is inferior to that of FDICA-PB, but the proposed combination technique can outperform both simple FDICA-PB and SIMO-ICA.Acoustics, Speech, and Signal Processing, 2004. Proceedings. (ICASSP '04). IEEE International Conference on; 06/2004