Single-Channel Multiple-Receiver Sound Source Localization System with Homomorphic Deconvolution and Linear Regression

Gaussian Process Regression for Single-Channel Sound Source Localization System Based on Homomorphic Deconvolution

To extract the phase information from multiple receivers, the conventional sound source localization system involves substantial complexity in software and hardware. Along with the algorithm complexity, the dedicated communication channel and individual analog-to-digital conversions prevent an increase in the system's capability due to feasibility. The previous study suggested and verified the single-channel sound source localization system, which aggregates the receivers on the single analog network for the single digital converter. This paper proposes the improved algorithm for the single-channel sound source localization system based on the Gaussian process regression with the novel feature extraction method. The proposed system consists of three computational stages: homomorphic deconvolution, feature extraction, and Gaussian process regression in cascade. The individual stages represent time delay extraction, data arrangement, and machine prediction, respectively. The optimal receiver configuration for the three-receiver structure is derived from the novel similarity matrix analysis based on the time delay pattern diversity. The simulations and experiments present precise predictions with proper model order and ensemble average length. The nonparametric method, with the rational quadratic kernel, shows consistent performance on trained angles. The Steiglitz-McBride model with the exponential kernel delivers the best predictions for trained and untrained angles with low bias and low variance in statistics.

Sound Source Localization Using a Convolutional Neural Network and Regression Model

In this research, a novel sound source localization model is introduced that integrates a convolutional neural network with a regression model (CNN-R) to estimate the sound source angle and distance based on the acoustic characteristics of the interaural phase difference (IPD). The IPD features of the sound signal are firstly extracted from time-frequency domain by short-time Fourier transform (STFT). Then, the IPD features map is fed to the CNN-R model as an image for sound source localization. The Pyroomacoustics platform and the multichannel impulse response database (MIRD) are used to generate both simulated and real room impulse response (RIR) datasets. The experimental results show that an average accuracy of 98.96% and 98.31% are achieved by the proposed CNN-R for angle and distance estimations in the simulation scenario at SNR = 30 dB and RT60 = 0.16 s, respectively. Moreover, in the real environment, the average accuracies of the angle and distance estimations are 99.85% and 99.38% at SNR = 30 dB and RT60 = 0.16 s, respectively. The performance obtained in both scenarios is superior to that of existing models, indicating the potential of the proposed CNN-R model for real-life applications.