Frequency band sensitivity analysis and sound source localization based on bionic hearing system

The study of simulating the human auditory system is an important approach to understanding the working mechanisms of auditory neurons and analyzing sound signal characteristics. This research focuses on simulating the functions of the auditory system by establishing models for auditory afferent pathways and the neural responses of the superior olivary complex. Additionally, the study aims to extract and analyze the spectral components of binaural signals to identify the frequency bands sensitive to acoustic cues of sound sources, such as interaural time difference (ITD) and interaural level difference (ILD). The model combines weighted Bayesian classification with the head-related transfer function (HRTF) to address the issue of missing auditory spatial perception cues in microphone array technology. This integration significantly improves the accuracy of single-point sound source localization and the trajectory tracking of moving sound sources in three-dimensional space. Experimental validation shows that the localization accuracy for a single-source sweeping signal can reach 96.89%, and the trajectory recognition accuracy for intermittent moving sound sources remains stable at over 80%. The research reveals the working mechanisms of the human auditory system and provides new solutions for sound source localization algorithms, spectral analysis, and feature extraction.