Integrated swarm intelligence and IoT for early and accurate remote voice-based pathology detection and water sound quality estimation

Abstract

In smart city design, artificial intelligence, health optimization, and the Internet of medical things are crucial in developing machine learning-based medical data analytics. The two main components of this approach are the Internet of Things (IoT) and swarm intelligence integration on the use of the Internet of Medical Things. The analysis of human speech and audio signals plays a crucial role which indicates physical and psychological well-being. Selecting appropriate features is very important and can greatly impact the overall signal processing and computation when developing these models based on speech applications. In the audio signal processing field, many features are available in the time, frequency, and statistical domain, and selecting proper features is a tedious task. But after feature selection, the stability analysis of these techniques is also equally important. This study considers these two problems by applying swarm intelligence-based feature selection techniques with higher stability. The optimized selected features are used for remote detection of voice-based pathological diseases, environmental sound detection in smart cities, acoustic sound quality assessment in amusement parks, and its impact on human psychological health. The proposed feature selection techniques are observed to perform comprehensively better than the standard speech features-based models in terms of both performance and computational complexities.