A face recognition algorithm based on the combine of image feature compensation and improved PSO

Since deep neural networks (DNN) has been introduced in face recognition field, many scholars have done lots of work to apply the technology into face recognition system. Local Binary Pattern (LBP)32 is another popular face recognition algorithm based on local features and is widely used in computer image recognition. Most of the face recognition algorithms above involve the setting of parameters, and the values of the parameters directly affect the recognition rate of the face recognition system. Zhang et al.51 proposed an Image Gradient Feature Compensation (IGFC) algorithm for face image recognition based on local feature compensation. Other studies have explored the use of particle swarm optimization (PSO) and image feature compensation techniques to improve face recognition technology.