Sound Source Distance Measurement Using Complex Sparse Bayesian Estimation with a Small Microphone Array System

This study aimed to measure the distance to a sound source using a small-scale microphone array through complex sparse Bayesian estimation. To achieve this, we extended the previous real Bayesian estimation, which relied solely on distance attenuation information, to a complex estimation that included phase information. Consequently, we demonstrated an expanded range of distance estimation for sinusoidal signals, while also highlighting the challenge of simultaneous distance estimation for two sound sources at the same frequency. When considering speech signals as the source, we showed that distance estimation remains possible for two sound sources owing to the sparsity of audio signals in the time-frequency domain.