Application of IAFSA-BP neural network in face orientation recognition

Abstract

In order to improve the recognition accuracy of BP neural network in face orientation recognition, an improved artificial fish swarm algorithm is proposed to optimize the weights and thresholds of BP neural network face orientation recognition model. The improved artificial fish swarm algorithm is based on the standard algorithm, introducing adaptive factors to make the horizon and step size adaptive change, and at the same time learning the solution before the result announcement, so as to improve the accuracy of the final result. Finally, the experimental results show that the effective combination of the improved artificial fish swarm algorithm and the BP neural network algorithm can improve the output accuracy of face orientation recognition of the BP neural network, and the running speed of the improved algorithm is significantly higher than that of the standard artificial fish swarm algorithm.