Here is a repository stores the classical sound source localization algorithms in the spherical domain, namely, PWD, DAS, SHMUSIC, SHMVDR, SHMLE. Here are two versions, Chinese and English
For the English readers, just open the notebook file which starts with "ENG".
For the Chinese readers, just open the notebook file which starts with "CHN".
Welcome to this tutorial, where we will delve into the technology of sound source localization in the spherical harmonic domain. First, we will introduce beamforming methods in detail, covering technologies such as PWD, DAS, and MVDR. Then, we will turn to subspace algorithms, exploring the frequency domain MUSIC and spherical harmonic domain SHMUSIC. In the final part of the tutorial, we will use the SHMLE method based on maximum likelihood estimation to estimate the Direction of Arrival (DoA) in a room. To comprehensively evaluate the performance of these algorithms, we have specially designed a series of experiments, examining the impact of different parameters such as broadband and narrowband signals, the number of microphones, the number of arrays, signal-to-noise ratio, and reverberation time on the results.
First, we will introduce several key library functions that we mainly use, then describe the problem and the simulation methods we adopt. We hope that by the end of this tutorial, you will have an understanding of sound source localization in the spherical harmonic domain.
The main contents of this notebook are as follows:
- Import Libraries and Modules
- What is Sound Source Localization
- Conducting Simulations
- Results and Conclusion
Some figures of results are shown below:
