Comparative analysis of KNN and CNN for Localization of Single Sound Source
Chaitali M. Gadre, Rashmika Patole, Shilpa P. Metkar · 2023
This paper proposes an integrated sound source localization through a Machine learning model such as KNN (K nearest neighbours algorithm) and CNN(Convolution neural network). Numerous industries, including robotics, surveillance, virtual reality, home entertainment, telecommunications, automotive systems, environmental monitoring, and speech processing, use sound source localization. Sound source localization is performed by using sound source angle and inter-phase difference (IPD) acoustic properties.and the accuracy obtained for KNN is 99.47% and for CNN is 95.24%. The comparison of the KNN and CNN is also presented in the paper.