Leveraging Bio-Acoustic Technologies for Advanced Sound Control Software Design in Commercial Applications

Zhiyong Chen
College of Information Engineering, Liaodong University, Dandong, Liaoning, 118000, China

Abstract:

This paper explores the utilization of bio-acoustic feature technology in enhancing sound control software design, emphasizing its application in bio-acoustic event detection and few-sample acoustic event recognition. By integrating bio-acoustic recognition technologies, this study enhances feature extraction processes using the MFCC-LPC method on the ARM platform, which improves the accuracy of parameter selection in acoustic recognition systems. The performance of three modeling methods—GMM, MFCC, and MFCC-LPC—was analyzed and compared to evaluate their effectiveness in voice control software. The results demonstrate a notable increase in recognition rates with extended speech durations, with a duration of 5 seconds adequately meeting practical application needs. Additionally, the system's capability for gender recognition was tested, analyzing 35-55 frames from voice samples, showing that it effectively operates within user-acceptable response times for speech lengths of 6-12 seconds. This indicates that the developed bio-acoustic feature-based system is particularly suitable for small to medium-sized user groups, offering significant implications for commercial biotechnology applications in sound control and interactive software solutions.