High Dimensional Problem Optimization Using Distributed Multi-agent PSO

Curse of dimensionality is a major difficulty with the classic optimization methods for high dimensional applications in which the problem size grows rapidly and mostly exponential with the number of space. In this work we present a simple yet effective multi-agent approach to apply distributed particle swarm optimization to meet such demand. Lip detection in color images, as a high-dimensional problem, has been investigated and a novel approach for obtaining an optimized lip-map was proposed. Experimental results show 92% correction rate which is 11% increase in comparison to the simple approach. A computational complexity analysis also shows the superiority of the proposed architecture to be used in other large scale application.