Cocktail parity problem solution based on modified blind extraction technique

Abstract

The Cocktail Party Problem solution is described as being the responsibility of isolating the voice signal in a noisy environment. Two popular methods for resolving this issue are blind source extraction and the Wiener filtering procedure. The blind source extraction approaches like fastICA, JADE and efficient fastICA are useful for extraction data from the mixed signals. The classical optimization techniques such as genetic algorithms or particle swarms for blind source extraction are mostly founded on the gradient and need the objective function, so the using of these techniques is very restricted. In the recent studies, the classical blind source separation techniques will not give the perfect solution for the cocktail parity problem. These methods struggle with convergence speed and accuracy issues as well. In order to enhance the separation process and get over these issues, this work adopted the glowworm swarm technique based on kurtosis as the objective function. The results show that the proposed technique produces good separation. The original and estimated signals are compared for similarity using the cross-correlation function.